In this tutorial, you will learn how to perform regression using Keras and Deep Learning. You will learn how to train a Keras neural network for regression and continuous value prediction, specifically in the context of house price prediction.

Today’s post kicks off a 3-part series on **deep learning, regression, and continuous value prediction.**

We’ll be studying Keras regression prediction in the context of house price prediction:

**Part 1:**Today we’ll be training a Keras neural network to**predict house prices based on categorical and numerical attributes**such as the number of bedrooms/bathrooms, square footage, zip code, etc.**Part 2:**Next week we’ll train a Keras**Convolutional Neural Network to predict house prices based on input images of the houses themselves**(i.e., frontal view of the house, bedroom, bathroom, and kitchen).**Part 3:**In two weeks we’ll define and train a neural network thatleading to better, more accurate house price prediction than the attributes or images alone.*combines*our categorical/numerical attributes with our images,

Unlike classification (which predicts * labels*), regression enables us to predict

*.*

**continuous values**For example, classification may be able to predict one of the following values: *{cheap, affordable, expensive}*.

Regression, on the other hand, will be able to predict an exact dollar amount, such as *“The estimated price of this house is $489,121”*.

In many real-world situations, such as house price prediction or stock market forecasting, applying regression rather than classification is *critical* to obtaining good predictions.

**To learn how to perform regression with Keras, just keep reading!**

Looking for the source code to this post?

Jump right to the downloads section.

## Regression with Keras

In the first part of this tutorial, we’ll briefly discuss the difference between classification and regression.

We’ll then explore the house prices dataset we’re using for this series of Keras regression tutorials.

From there, we’ll configure our development environment and review our project structure.

Along the way, we will learn how to use Pandas to load our house price dataset and define a neural network that for Keras regression prediction.

Finally, we’ll train our Keras network and then evaluate the regression results.

### Classification vs. Regression

Typically on the PyImageSearch blog, we discuss Keras and deep learning in the context of classification — predicting a label to characterize the contents of an image or an input set of data.

**Regression**, on the other hand, enables us to predict continuous values. Let’s again consider the task of house price prediction.

As we know, classification is used to predict a class label.

For house price prediction we may define our categorical labels as:

1 |
labels = {very cheap, cheap, affordable, expensive, very expensive} |

If we performed classification, our model could then learn to predict one of those five values based on a set of input features.

However, those labels are just that — categories that represent a potential range of prices for the house but do *nothing* to represent the *actual cost* of the home.

**In order to predict the actual cost of a home, we need to perform regression.**

Using regression we can train a model to predict a continuous value.

For example, while classification may only be able to predict a label, regression could say:

*“Based on my input data, I estimate the cost of this house to be $781,993.”*

**Figure 1** above provides a visualization of performing both classification and regression.

In the rest of this tutorial, you’ll learn how to train a neural network for regression using Keras.

### The House Prices Dataset

The dataset we’ll be using today is from 2016 paper, *House price estimation from visual and textual features*, by Ahmed and Moustafa.

**The dataset includes both numerical/categorical attributes along with images for 535 data points,** making it and excellent dataset to study for regression and mixed data prediction.

The house dataset includes **four numerical and categorical attributes:**

- Number of bedrooms
- Number of bathrooms
- Area (i.e., square footage)
- Zip code

These attributes are stored on disk in CSV format.

We’ll be loading these attributes from disk later in this tutorial using pandas , a popular Python package used for data analysis.

**A total of four images are also provided for each house:**

- Bedroom
- Bathroom
- Kitchen
- Frontal view of the house

**The end goal of the houses dataset is to predict the price of the home itself.**

In today’s tutorial, we’ll be working with *just* the numerical and categorical data.

Next week’s blog post will discuss working with the image data.

And finally, two weeks from now we’ll *combine* the numerical/categorical data with the images to obtain our best performing model.

But before we can train our Keras model for regression, we first need to configure our development environment and grab the data.

### Configuring Your Development Environment

For this 3-part series of blog posts, you’ll need to have the following packages installed:

**NumPy****scikit-learn****pandas****Keras**with the**TensorFlow**backend (CPU or GPU)**OpenCV**(for the next two blog posts in the series)

Luckily most of these are easily installed with pip, a Python package manager.

Let’s install the packages now, ideally into a virtual environment as shown (you’ll need to create the environment):

1 2 3 4 5 6 |
$ workon house_prices $ pip install numpy $ pip install scikit-learn $ pip install pandas $ pip install tensorflow # or tensorflow-gpu $ pip install keras |

Notice that I haven’t instructed you to install OpenCV yet. The OpenCV install can be slightly involved — especially if you are compiling from source. Let’s look at our options:

**Compiling from source**gives us the full install of OpenCV and provides access to optimizations, patented algorithms, custom software integrations, and more. The good news is that all of my**OpenCV install tutorials**are meticulously put together and updated regularly. With patience and attention to detail, you can compile from source just like I and many of my readers do.**Using pip to install OpenCV**is hands-down the fastest and easiest way to get started with OpenCV and essentially just checks prerequisites and places a precompiled binary that will work on most systems into your virtual environment site-packages. Optimizations may or may not be active. The big caveat is that the maintainer has elected not to include patented algorithms for fear of lawsuits. There’s nothing wrong with using patented algorithms for educational and research purposes, but you should use alternative algorithms commercially. Nevertheless, the pip method is a great option for beginners just remember that you don’t have the full install.

Pip is sufficient for this 3-part series of blog posts. You can install OpenCV in your environment via:

1 2 |
$ workon house_prices $ pip install opencv-contrib-python |

Please reach out to me if you have any difficulties getting your environment established.

### Downloading the House Prices Dataset

Before you download the dataset, go ahead and grab the source code to this post by using * “Downloads” *section.

From there, unzip the file and navigate into the directory:

1 2 3 |
$ cd path/to/downloaded/zip $ unzip keras-regression.zip $ cd keras-regression |

From there, you can download the House Prices Dataset using the following command:

1 |
$ git clone https://github.com/emanhamed/Houses-dataset |

When we are ready to train our Keras regression network you’ll then need to supply the path to the Houses-dataset directory via command line argument.

### Project structure

Now that you have the dataset, go ahead and use the tree command with the same arguments shown below to print a directory + file listing for the project:

1 2 3 4 5 6 7 8 9 10 11 12 |
$ tree --dirsfirst --filelimit 10 . ├── Houses-dataset │ ├── Houses Dataset [2141 entries] │ └── README.md ├── pyimagesearch │ ├── __init__.py │ ├── datasets.py │ └── models.py └── mlp_regression.py 3 directories, 5 files |

The dataset downloaded from GitHub now resides in the Houses-dataset/ folder.

The
pyimagesearch/ directory is actually a module included with the code * “Downloads”* where inside, you’ll find:

- datasets.py : Our script for loading the numerical/categorical data from the dataset
- models.py : Our Multi-Layer Perceptron architecture implementation

These two scripts will be reviewed today. Additionally, we’ll be reusing both datasets.py and models.py (with modifications) in the next two tutorials to keep our code organized and reusable.

The regression + Keras script is contained in mlp_regression.py which we’ll be reviewing it as well.

### Loading the House Prices Dataset

Before we can train our Keras regression model we first need to load the numerical and categorical data for the houses dataset.

Open up the datasets.py file an insert the following code:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 |
# import the necessary packages from sklearn.preprocessing import LabelBinarizer from sklearn.preprocessing import MinMaxScaler import pandas as pd import numpy as np import glob import cv2 import os def load_house_attributes(inputPath): # initialize the list of column names in the CSV file and then # load it using Pandas cols = ["bedrooms", "bathrooms", "area", "zipcode", "price"] df = pd.read_csv(inputPath, sep=" ", header=None, names=cols) |

We begin by importing libraries and modules from scikit-learn, pandas, NumPy and OpenCV. OpenCV will be used next week as we’ll be adding the ability to load images to this script.

On **Line 10**, we define the
load_house_attributes function which accepts the path to the input dataset.

Inside the function we start off by defining the names of the columns in the CSV file (**Line 13**). From there, we use pandas’ function,
read_csv to load the CSV file into memory as a date frame (
df ) on **Line 14**.

Below you can see an example of our input data, including the number of bedrooms, number of bathrooms, area (i.e., square footage), zip code, code, and finally the target price our model should be trained to predict:

1 2 3 4 5 6 |
bedrooms bathrooms area zipcode price 0 4 4.0 4053 85255 869500.0 1 4 3.0 3343 36372 865200.0 2 3 4.0 3923 85266 889000.0 3 5 5.0 4022 85262 910000.0 4 3 4.0 4116 85266 971226.0 |

Let’s finish up the rest of the load_house_attributes function:

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 |
# determine (1) the unique zip codes and (2) the number of data # points with each zip code zipcodes = df["zipcode"].value_counts().keys().tolist() counts = df["zipcode"].value_counts().tolist() # loop over each of the unique zip codes and their corresponding # count for (zipcode, count) in zip(zipcodes, counts): # the zip code counts for our housing dataset is *extremely* # unbalanced (some only having 1 or 2 houses per zip code) # so let's sanitize our data by removing any houses with less # than 25 houses per zip code if count < 25: idxs = df[df["zipcode"] == zipcode].index df.drop(idxs, inplace=True) # return the data frame return df |

In the remaining lines, we:

- Determine the unique set of zip codes and then count the number of data points with each unique zip code (
**Lines 18 and 19**). - Filter out zip codes with low counts (
**Line 28**). For some zip codes we only have one or two data points, making it extremely challenging, if not impossible, to obtain accurate house price estimates. - Return the data frame to the calling function (
**Line 33**).

Now let’s create the process_house_attributes function used to preprocess our data:

35 36 37 38 39 40 41 42 43 |
def process_house_attributes(df, train, test): # initialize the column names of the continuous data continuous = ["bedrooms", "bathrooms", "area"] # performin min-max scaling each continuous feature column to # the range [0, 1] cs = MinMaxScaler() trainContinuous = cs.fit_transform(train[continuous]) testContinuous = cs.transform(test[continuous]) |

We define the function on **Line 35**. The
process_house_attributes function accepts three parameters:

- df : Our data frame generated by pandas (the previous function helps us to drop some records from the data frame)
- train : Our training data for the House Prices Dataset
- test : Our testing data.

Then on **Line 37**, we define the columns of our our continuous data, including bedrooms, bathrooms, and size of the home.

We’ll take these values and use scikit-learn’s
MinMaxScaler to scale the continuous features to the range *[0, 1]* (**Lines 41-43**).

Now we need to pre-process our categorical features, namely the zip code:

45 46 47 48 49 50 51 52 53 54 55 56 57 |
# one-hot encode the zip code categorical data (by definition of # one-hot encoing, all output features are now in the range [0, 1]) zipBinarizer = LabelBinarizer().fit(df["zipcode"]) trainCategorical = zipBinarizer.transform(train["zipcode"]) testCategorical = zipBinarizer.transform(test["zipcode"]) # construct our training and testing data points by concatenating # the categorical features with the continuous features trainX = np.hstack([trainCategorical, trainContinuous]) testX = np.hstack([testCategorical, testContinuous]) # return the concatenated training and testing data return (trainX, testX) |

First, we’ll one-hot encode the zip codes (**Lines 47-49**).

Then we’ll concatenate the *categorical features* with the *continuous features* using NumPy’s
hstack function (**Lines 53 and 54**), returning the resulting training and testing sets as a tuple (**Line 57**).

Keep in mind that now *both* our categorical features and continuous features are *all* in the range *[0, 1].*

### Implementing a Neural Network for Regression

Before we can train a Keras network for regression, we first need to define the architecture itself.

Today we’ll be using a simple Multilayer Perceptron (MLP) as shown in **Figure 5**.

Open up the models.py file and insert the following code:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 |
# import the necessary packages from keras.models import Sequential from keras.layers.normalization import BatchNormalization from keras.layers.convolutional import Conv2D from keras.layers.convolutional import MaxPooling2D from keras.layers.core import Activation from keras.layers.core import Dropout from keras.layers.core import Dense from keras.layers import Flatten from keras.layers import Input from keras.models import Model def create_mlp(dim, regress=False): # define our MLP network model = Sequential() model.add(Dense(8, input_dim=dim, activation="relu")) model.add(Dense(4, activation="relu")) # check to see if the regression node should be added if regress: model.add(Dense(1, activation="linear")) # return our model return model |

First, we’ll import all of the necessary modules from Keras (**Lines 2-11**). We’ll be adding a Convolutional Neural Network to this file in *next week’s tutorial*, hence the additional imports that aren’t utilized here today.

Let’s define the MLP architecture by writing a function to generate it called create_mlp .

The function accepts two parameters:

- dim : Defines our input dimensions
- regress : A boolean defining whether or not our regression neuron should be added

We’ll go ahead and start construction our MLP with a
dim-8-4 architecture (**Lines 15-17**).

If we are performing regression, we add a
Dense layer containing a single neuron with a linear activation function (**Lines 20 and 21**). Typically we use ReLU-based activations, but since we are performing regression we need a linear activation.

Finally, our
model is returned on **Line 24**.

### Implementing our Keras Regression Script

It’s now time to put all the pieces together!

Open up the mlp_regression.py file and insert the following code:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
# import the necessary packages from keras.optimizers import Adam from sklearn.model_selection import train_test_split from pyimagesearch import datasets from pyimagesearch import models import numpy as np import argparse import locale import os # construct the argument parser and parse the arguments ap = argparse.ArgumentParser() ap.add_argument("-d", "--dataset", type=str, required=True, help="path to input dataset of house images") args = vars(ap.parse_args()) |

We begin by importing necessary packages, modules, and libraries.

Namely, we’ll need the Adam optimizer from Keras, train_test_split from scikit-learn, and our datasets + models functions from the pyimagesearch module.

Additionally, we’ll use math features from NumPy for collecting statistics when we evaluate our model.

The argparse module is for parsing command line arguments.

Our script requires just one command line argument
--dataset (**Lines 12-15**). You’ll need to provide the
--dataset switch and the actual path to the dataset when you go to run the training script in your terminal.

Let’s load the house dataset attributes and construct our training and testing splits:

17 18 19 20 21 22 23 24 25 26 |
# construct the path to the input .txt file that contains information # on each house in the dataset and then load the dataset print("[INFO] loading house attributes...") inputPath = os.path.sep.join([args["dataset"], "HousesInfo.txt"]) df = datasets.load_house_attributes(inputPath) # construct a training and testing split with 75% of the data used # for training and the remaining 25% for evaluation print("[INFO] constructing training/testing split...") (train, test) = train_test_split(df, test_size=0.25, random_state=42) |

Using our handy
load_house_attributes function, and by passing the
inputPath to the dataset itself, our data is loaded into memory (**Lines 20 and 21**).

Our training (75%) and testing (25%) data is constructed via **Line 26** and scikit-learn’s
train_test_split method.

Let’s scale our house pricing data:

28 29 30 31 32 33 |
# find the largest house price in the training set and use it to # scale our house prices to the range [0, 1] (this will lead to # better training and convergence) maxPrice = train["price"].max() trainY = train["price"] / maxPrice testY = test["price"] / maxPrice |

As stated in the comment, scaling our house prices to the range *[0, 1]* will allow our model to more easily train and converge. Scaling the output targets to *[0, 1]* will reduce the range of our output predictions (versus *[0,
maxPrice ]*) and make it not only * easier and faster to train our network* but enable our model to

**obtain better results as well.**Thus, we grab the maximum price in the training set (**Line 31**), and proceed to scale our training and testing data accordingly (**Lines 32 and 33**).

Let’s process the house attributes now:

16 17 18 19 20 |
# process the house attributes data by performing min-max scaling # on continuous features, one-hot encoding on categorical features, # and then finally concatenating them together print("[INFO] processing data...") (trainX, testX) = datasets.process_house_attributes(df, train, test) |

Recall from the datasets.py script that the process_house_attributes function:

- Pre-processes our categorical and continuous features.
- Scales our continuous features to the range
*[0, 1]*via min-max scaling. - One-hot encodes our categorical features.
- Concatenates the categorical and continuous features to form the final feature vector.

Now let’s go ahead and fit our MLP model to the data:

41 42 43 44 45 46 47 48 49 50 51 52 |
# create our MLP and then compile the model using mean absolute # percentage error as our loss, implying that we seek to minimize # the absolute percentage difference between our price *predictions* # and the *actual prices* model = models.create_mlp(trainX.shape[1], regress=True) opt = Adam(lr=1e-3, decay=1e-3 / 200) model.compile(loss="mean_absolute_percentage_error", optimizer=opt) # train the model print("[INFO] training model...") model.fit(trainX, trainY, validation_data=(testX, testY), epochs=200, batch_size=8) |

Our
model is initialized with the
Adam optimizer (**Lines 45 and 46**) and then compiled (**Line 47**). Notice that we’re using *mean absolute percentage error* as our loss function, indicating that we seek to minimize the *mean percentage difference* between the predicted price and the actual price.

The actual training process is kicked off on **Lines 51 and 52.**

After training is complete we can evaluate our model and summarize our results:

54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 |
# make predictions on the testing data print("[INFO] predicting house prices...") preds = model.predict(testX) # compute the difference between the *predicted* house prices and the # *actual* house prices, then compute the percentage difference and # the absolute percentage difference diff = preds.flatten() - testY percentDiff = (diff / testY) * 100 absPercentDiff = np.abs(percentDiff) # compute the mean and standard deviation of the absolute percentage # difference mean = np.mean(absPercentDiff) std = np.std(absPercentDiff) # finally, show some statistics on our model locale.setlocale(locale.LC_ALL, "en_US.UTF-8") print("[INFO] avg. house price: {}, std house price: {}".format( locale.currency(df["price"].mean(), grouping=True), locale.currency(df["price"].std(), grouping=True))) print("[INFO] mean: {:.2f}%, std: {:.2f}%".format(mean, std)) |

**Line 56** instructs Keras to make predictions on our testing set.

Using the predictions, we compute the:

- Difference between
*predicted*house prices and the*actual*house prices (**Line 61**). - Percentage difference (
**Line 62**). - Absolute percentage difference (
**Line 63**).

From there, on **Lines 67 and 68**, we calculate the mean and standard deviation of the absolute percentage difference.

The results are printed via **Lines 72-75**.

Regression with Keras wasn’t so tough, now was it?

Let’s train the model and analyze the results!

### Keras Regression Results

To train our own Keras network for regression and house price prediction make sure you have:

- Configured your development environment according to the guidance above.
- Used the
section of this tutorial to download the source code.**“Downloads”** - Downloaded the house prices dataset based on the instructions in the
*“The House Prices Dataset”*section above.

From there, open up a terminal and supply the following command (making sure the --dataset command line argument points to where you downloaded the house prices dataset):

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 |
$ python mlp_regression.py --dataset Houses-dataset/Houses\ Dataset/ [INFO] loading house attributes... [INFO] constructing training/testing split... [INFO] processing data... [INFO] training model... Train on 271 samples, validate on 91 samples Epoch 1/200 271/271 [==============================] - 0s 680us/step - loss: 84.0388 - val_loss: 61.7484 Epoch 2/200 271/271 [==============================] - 0s 110us/step - loss: 49.6822 - val_loss: 50.4747 Epoch 3/200 271/271 [==============================] - 0s 112us/step - loss: 42.8826 - val_loss: 43.5433 Epoch 4/200 271/271 [==============================] - 0s 112us/step - loss: 38.8050 - val_loss: 40.4323 Epoch 5/200 271/271 [==============================] - 0s 112us/step - loss: 36.4507 - val_loss: 37.1915 Epoch 6/200 271/271 [==============================] - 0s 112us/step - loss: 34.3506 - val_loss: 35.5639 Epoch 7/200 271/271 [==============================] - 0s 111us/step - loss: 33.2662 - val_loss: 37.5819 Epoch 8/200 271/271 [==============================] - 0s 108us/step - loss: 32.8633 - val_loss: 30.9948 Epoch 9/200 271/271 [==============================] - 0s 110us/step - loss: 30.4942 - val_loss: 30.6644 Epoch 10/200 271/271 [==============================] - 0s 107us/step - loss: 28.9909 - val_loss: 28.8961 ... Epoch 195/200 271/271 [==============================] - 0s 111us/step - loss: 20.8431 - val_loss: 21.4466 Epoch 196/200 271/271 [==============================] - 0s 109us/step - loss: 22.2301 - val_loss: 21.8503 Epoch 197/200 271/271 [==============================] - 0s 112us/step - loss: 20.5079 - val_loss: 21.5884 Epoch 198/200 271/271 [==============================] - 0s 108us/step - loss: 21.0525 - val_loss: 21.5993 Epoch 199/200 271/271 [==============================] - 0s 112us/step - loss: 20.4717 - val_loss: 23.7256 Epoch 200/200 271/271 [==============================] - 0s 107us/step - loss: 21.7630 - val_loss: 26.0129 [INFO] predicting house prices... [INFO] avg. house price: $533,388.27, std house price: $493,403.08 [INFO] mean: 26.01%, std: 18.11% |

As you can see from our output, our initial mean absolute percentage error starts off as high as 84% and then quickly drops to under 30%.

By the time we finish training we can see our network starting to overfit a bit. Our training loss is as low as ~21%; however, our validation loss is at ~26%.

Computing our final mean absolute percentage error we obtain a final value of **26.01%.**

**What does this value mean?**

Our final mean absolute percentage error implies, that on average, our network will be ~26% off in its house price predictions with a standard deviation of ~18%.

### Limitations of the House Price Dataset

Being 26% off in a house price prediction is a good start but is certainly *not* the type of accuracy we are looking for.

That said, this prediction accuracy can also be seen as a *limitation* of the house price dataset itself.

Keep in mind that the dataset only includes four attributes:

- Number of bedrooms
- Number of bathrooms
- Area (i.e., square footage)
- Zip code

Most other house price datasets include *many* more attributes.

For example, the Boston House Prices Dataset includes a total of *fourteen attributes* which can be leveraged for house price prediction (although that dataset does have some racial discrimination).

The Ames House Dataset *includes over 79 different attributes* which can be used to train regression models.

When you think about it, the fact that we are able to even obtain 26% mean absolute percentage error without the knowledge of an expert real estate agent is fairly reasonable given:

- There are only 535 total houses in the dataset (we only used
**362 total houses**for the purpose of this guide). - We only have
**four attributes**to train our regression model on. - The attributes themselves, while important in describing the home itself,
**do little to characterize the area surrounding the house.** **The house prices are incredibly varied**with a mean of $533K and a standard deviation of $493K (based on our filtered dataset of 362 homes).

**With all that said, learning how to perform regression with Keras is an important skill!**

In the next two posts in this series I’ll be showing you how to:

- Leverage the images provided with the house price dataset to train a CNN on them.
- Combine our numerical/categorical data with the house images, leading to a model that outperforms all of our previous Keras regression experiments.

## Summary

In this tutorial, you learned how to use the Keras deep learning library for regression.

Specifically, we used Keras and regression to predict the price of houses based on four numerical and categorical attributes:

- Number of bedrooms
- Number of bathrooms
- Area (i.e., square footage)
- Zip code

Overall our neural network obtained a mean absolute percentage error of 26.01%, implying that, on average, our house price predictions will be off by 26.01%.

That raises the questions:

- How can we
*better*our house price prediction accuracy? - What if we
*leveraged images*for each house? Would that improve accuracy? - Is there some way to
*combine*both our categorical/numerical attributes with our image data?

To answer these questions you’ll need to stay tuned for the remaining to tutorials in this Keras regression series.

**To download the source code to this post (and be notified when the next tutorial is published here on PyImageSearch), just enter your email address in the form below.**

Thanks for the great article on regression! One question I have is what rule do you use to determine the number of layers and neurons per layer? Is it a function of the number of inputs?

Thanks!

It’s a hyperparameter that you tune. A general rule of thumb for multi-layer perceptrons, like the one covered here, is to reduce the number of neurons per layer. Sometimes you may dramatically reduce the number of nodes, other times the network will be deeper and the number of neurons will gradually reduce. It’s very much a set of parameters, called hyperparameters, you need to tune. I cover my best practices for defining and training neural networks inside Deep Learning for Computer Vision with Python just in case you are interested in learning more.

<snip Unlike classification (which predicts labels), regression enables us to predict continuous values.

For example, classification may be able to predict one of the following values: {cheap, affordable, expensive}.

Regression, on the other hand, will be able to predict an exact dollar amount, such as “The estimated price of this house is $489,121”.

PL] Dear Adrian, You stand alone in your explanation!! Absolutely magnificent! Thanks once again and I look forward to the next post.

Thanks Pranav 🙂

Adrian — Thanks for the interesting blog!

It seems that you touch upon a little bit just how challenging it can be to build a representative dataset for training and evaluation. It is very impressive to me just how good the results can be from such a small training set.

Thanks David, I’m glad you liked the post!

Have you considered replacing zip code with latitude and longitude values? Essentially converting hundreds or thousands of sparse, one-hot encoded values into two continuous values?

I’m curious to see if performance degrades or improves. I’ve seen improvements when using tree methods with this transformation since the model can group geographic similarity more naturally than with categories which don’t encode similarity.

Technically there’s no reason why you couldn’t do that but I’m not convinced it would improve prediction accuracy in a material way. The area/zip code of where a house is often relates significantly to the price of the house. Replacing that with latitude and longitude may or may not improve accuracy, it’s hard to tell without running the experiment.

zio code is likely to tell whether you are in a rich or poor region in the USA. (and results cannot be generalized to other parts of the world). Using lat/long would lead users feel this regression would work in Europe, say…

Hi Adrian,

Will u be using the meras functional api to combine the cnn model w the model from this post?

Thanks

Yes, you are correct.

Thanks for the great post. It would be great if you could please provide the code/tutiroal of same prediction using pytorch.

Very good tutorial about regression. Will be a stepping stone for beginners in Machine learning.

Thanks Adrian.

Thanks Dipin!

Hi Adrian,

Now that i have the model created how can i predict a new house value? do i need to re-train it every time? what part of code can i use to just predict a new value?

Thanks

You don’t need to retrain the model each time, you can just save and load your Keras model from disk. The

`model.predict`

function can be used to predict new home prices based on your input features.Hi Adrian,

Really detailed guide, thank you so much for making it, i had one question, how do i handle a scenario where i have multiple categorical columns. In your example Zip is your only categorical column, i’m trying to apply this to my own data and have LoanType and Zip as mine, passing these as an array to LabelBinarizer throws a ValueError: Multioutput target data is not supported with label binarization.

I was wondering if there’s something simple i’m messing up?

You would need to create a LabelBinarizer for each of your categorical columns and then concatenate the output of them. You could look into using scikit-learn’s MultiLabelBinarizer as well.

Thanks Adrian, managed to get it working by doing that, really great tutorial, can’t wait for the convnet one. I’ve trained my model and it all seems to work OK, i’m now trying to predict on new data it’s never seen before (not in the train/test sets) and the shape of the NumpyArray that creates is different (The one i trained it on had 149588,425, the days worth of data i’m now trying to predict is 514, 137 due to a variability in a set’s amount of zipcodes).

This is probably a dumb question but am i missing a step?

It’s hard to say without seeing your data but my guess is that you haven’t pre-processed your testing data in the same manner as your testing data. Your label encoder/transformer was likely created on your training data and then due to a logic error was reinstantiated and re-created on your testing data.

If your data is skewed and your training/testing data contains values not in the other you can either:

1. Apply the transformer to all the data before the split (not technically correct if you want to publish a paper but it will get you a proof of concept)

2. Try to apply missing numbers or interpolation into the transformer process

Hi Adrian,

When I plotted the data I saw two potential outliers. Removing them improved the average result. By the way, great post!

Best,

Thanks for sharing Guilherme! What were the two outlier data points and how much did the results improve?

The two outliers were houses that had the price over 3 million dollars. One of them was removed when you filtered the zip code. Over 100 experiments the result with those outliers was “mean: 22.763, std: 21.950”. Without them the result was “mean: 21.922, std: 20.338”. Considering we only had four features, is this a significant improvement?

That’s certainly better but you would want to run a 5-fold or 10-fold cross-validation experiment with and without the outliers to confirm.

Hi Adrian, Thanks once again for nice article. Just wanted to know can I do all this or Deep learning coding practice docker installated on windows 10 rather than using Ubuntu through virtual machines(VMware). ?

Hey Vikas — I would recommend you spend some time reading up on Docker and practice installing an Ubuntu Docker instance on your Windows machine. Once you have Docker up and running you can follow any of my install guides to get up and running.

Hi Adrian

What is the difference between LabelBinarizer and OneHotEncoder (both are in sklearn.preprocessing) ?

Thank you.

OneHotEncoder assumes that your data is already in integer format. The LabelBinarizer doesn’t care and will first encode as integers (if you input strings as labels) and then will perform the one-hot encoding.

Hi Adrian

Just curious why you did not scale the prediction with maxPrice to get the dollar value of the output.

preds.flatten() * maxPrice

Thank you.

It really doesn’t matter, it’s just scaling. Both the testing and training data have been scaled by maxPrice. You could rescale them but if you wanted to obtain the raw dollar amount but that won’t affect the final computed percentage error.

Hi Adrian,

I have done several deep learning projects doing regression with Keras but I’ve always used a ReLu activation for the final neuron. Is there any advantage to using a linear activation? I thought that ReLu would be better because it prevents predictions from being negative.

But what happens if you are working on a problem where your network

shouldpredict negative values? If you place a ReLU at the end you’ll never be able to predict those negative values. Use a linear activation for your final output for regression — your network should be stable enough to predict the values you want. If not, your network architecture or training procedure should be updated.Hi Adrian, nice tutorial.

One question – using your code, how can we predict the price of one house, that is not on the training/test set?

Regards.

Keep in mind that a model is only as good as the data it was trained on. The model used here today would be capable of predicting models from the same data distribution as the training/testing set. You would need to have the four values required by our model:

1. # of bedrooms

2. # of bathrooms

3. Area

4. Zip code

From there you would pass those into the model and obtain your prediction.

If you are in a country whithout zip code, you cannot

(this dataset is not meant to be generalized : it is meant to train you)

Hello,Adrian can you please make a tutorial on how to detect vehicles and measure their speed using Raspberry pi.

I’ll actually be covering that exact project in my upcoming Computer Vision + Raspberry Pi book, stay tuned!

This tutorial worked great on a RPi : if timing is exact, RPi trains 8 times slower than Adrian PC.

There was a tiny minor flow: line 74 of mlp-regression has to be removed/ commented out .

Thanks for sharing, Denis! I wouldn’t recommend actually training the networks themselves on the Pi as it is resource constrained. Typically the pipeline I suggest is:

1. Train your network on your laptop, desktop, or deep learning rig

2. Export the model

3. Transfer the model to the Pi

4. Perform prediction/inference on the Pi

Thanks again for sharing the benchmark though!

Can’t wait for the next part of this series 😀

Thanks Allan!

Hi Adrian,

I had a simple query. After we have trained our model, I saved it. And now I want to predict real values for houses. Example for input as 4 4 4053 85255, I want to predict the house price say 869500. I think it can be done using model.predict() by providing scaled input features but was not sure how to do it exactly.

For the continuous features you’ll want to use the “cs” MinMaxScaler and call the

`.transform`

method. Same goes for the categorical features but this time using the`.transform`

method of the LabelBinarizer. Concatenate them just like we do in the post and then pass them into`model.predict`

.Give it a try and spend time hacking with the code. Gain hands on experience and fight with the code if you need to, it’s one of the best ways to learn!

i have the same question, but i dont know how to solve 🙁

thanks for this awesome tutorial

hey apoorva can you please tell me on how to and where to call predict.model() method nd where to pass the features like the no.of bathrooms and all

Hey Adrian!

Thanks for the great tutorial man, it was excellent. Quite intuitive and retrospective at the same time, something that everyone learning machine learning looks for.

As a question pal, I’d like to ask something I’ve been researching on for quite a while now. .Since you’ve used Convolutional Neural Networks in this area, can we, with sheer luck, use the same logic for automatic captcha recognition and possibly, answering too?

I ask since we can even play games using gestures and this concept being quite interesting research topic.

This thing is so interesting it’s been eating my head for quite a while now.

Great question. My friend Adam Geitgey used the code from Deep Learning for Computer Vision with Python as a staring point and then built his own captcha breaker. I would suggest starting there.

Hi, Adrian.

Thank you for this posting.

CSV to scalar-value regression is good… but,

I’m interesting that “Image to scalar-value regression”.

Bounding box regression in R-CNN is also an image2value regression.

At next time, What about deal with this trial?

I hope ~

Thanks you.

Hey Gromit — I actually cover Faster R-CNNs inside my book, Deep Learning for Computer Vision with Python. If you’re interested in R-CNNs I would suggest starting there. I also covered image-to-value regression in this tutorial.

Excellent. Thanks Adrain.

You are welcome!

in the loadhouseattribute when we pass the “inputpath” parameter does it mean we have to pass the path of the directory storing inputdataset or we just have to write the “inputpath” as it is?

You pass in the path to the directory that contains the actual Houses dataset (the directory with the CSV file and images).

Hi, great post. I wonder, is it truly necessary to use MinMaxScaler? Does it help the model to converge? Thank you.

Yes, it most certainly does. Try experimenting for yourself. What happens if you remove the MinMaxScaling?

Hi Adrian…I m getting error in jupyter notebook

ModuleNotFoundError: No module named ‘pyimagesearch’…please suggest ..thanks..

You need to use the “Downloads” section of the blog post to download the source code. That download contains the “pyimagesearch” module.

Very good tutorial mate, Thanks!

Thanks, I’m gad you enjoyed it!

how to find the house price for just 1 house attributes

I think you mean 1 house vector of features, right? Meaning if you a the house image + categorical attributes + numerical attributes? If so, call “model.predict” and pass in the features.

Hi Adrian,

Thank you so much for your wonderful tutorials! I’ve learned so much! You’ve made a difficult topic easy to understand. That is truly an art!

In your 3-part series, may I suggest doing internal links to the other parts in the series? It’ll be easier for readers to find the other parts of the series and I think this will also help with SEO.

Thanks for the suggestion.

Hi Adrian.

I love your tutorials! They are very informative. I am curious about GANs, specifically for regression. Can you post an entry about this topic and compare the accuracy to that of normal regression?

Thanks for the suggestion. I will consider it but cannot guarantee if/when I will cover it.

Hi Adrian,

Thanks for the tutorial, I have doubt in features normalization, why you are converting zip codes to one hot encode only? Even we can also areas into one hot encode also? right?

Hi Adrian,

great tutorials ..I had a question.How can we check which features have the highest coefficients, after doing regression?(similar to what we do in machine learning)

Hi Adrian,

thanks for this detailled work.

You scaled the output values using the maximum of the prices in the training set.

Is there a specific reason why you did not use the maximum of ALL prices, i.e. contained in both the training and the testing set?

Actually, I scaled the prices of ALL data points but I found the maximum value from the training data only. You are not allowed to glean information from the testing set — it’s used only for testing, therefore the maximum must be computed from the training data.