Washington-Bike-Demand-Prediction

KO | EN

gitlite — search

#python #react #typescript #llm #golang #go #rust #nodejs #java #windows #cli #vue #security #openai

Washington-Bike-Demand-Prediction

(★ 9)

Using different machine learning and deep learning techniques, the demand for Washington Rental Bikes would be predicted so that the companies could ensure that people receive bikes as and when needed.

Washington-Bike-Demand-Prediction 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# 🚴‍♀️ Washington Bike Demand Prediction

[![](https://img.shields.io/badge/Python-FFD43B?style=for-the-badge&logo=python&logoColor=darkgreen)](https://www.python.org)  [![](https://img.shields.io/badge/TensorFlow-FF6F00?style=for-the-badge&logo=TensorFlow&logoColor=white)](https://www.tensorflow.org) [![](https://img.shields.io/badge/scikit_learn-F7931E?style=for-the-badge&logo=scikit-learn&logoColor=white)](https://scikit-learn.org/stable/) [![](https://img.shields.io/badge/SciPy-654FF0?style=for-the-badge&logo=SciPy&logoColor=white)](https://www.scipy.org) [![](https://img.shields.io/badge/Numpy-777BB4?style=for-the-badge&logo=numpy&logoColor=white)](https://numpy.org) [![](https://img.shields.io/badge/Pandas-2C2D72?style=for-the-badge&logo=pandas&logoColor=white)](https://pandas.pydata.org)  [![](https://img.shields.io/badge/Plotly-239120?style=for-the-badge&logo=plotly&logoColor=white)](https://plotly.com) [![](https://img.shields.io/badge/Keras-D00000?style=for-the-badge&logo=Keras&logoColor=white)](https://keras.io) [![](https://img.shields.io/badge/conda-342B029.svg?&style=for-the-badge&logo=anaconda&logoColor=white)](https://www.anaconda.com)

## Introduction

There are different bike companies such as [__Zagster__](https://en.wikipedia.org/wiki/Zagster) and [__Lime Bikes__](https://www.li.me/en-US/home) present all over the __US__ to ensure that people get rides when needed. As a result, there is an increase in demand for these bikes as some people move to these __cost-efficient__ and __environment-friendly__ transport options. As a result of this transition, there is variation in the demand for bikes in different regions.

## Challenges

One of the challenges that these companies face is to know the number of bikes that must be placed at different locations at different instances of time to ensure that they get maximum profit and give the people their rides. Sometimes there is a possibility of people missing out on these bikes due to their unavailability. On the contrary, there are also instances when the demand for these bikes is low while they are highly available in many locations without being used. Therefore, it becomes important to tackle these instances and understand the demand for bikes for different days and scenarios.

## Data Science and Machine Learning

With the help of __machine learning__ and __deep learning__, this problem could be addressed and this would ensure that the demand for the bikes is known beforehand thus, the companies could ensure that there are __adequate bikes__ present in different locations.

## Exploratory Data Analysis (EDA)

Therefore, with the help of __data visualization__ and __machine learning__, bike rental companies would be able to understand the total number of bikes that must be present at different instances of time and thus, they would be able to predict the demand for the bikes in the future. This would ensure that the companies save millions of dollars by giving the right service to different people who are in need.

## Metrics

Since this is a __regression__ problem, we consider metrics that take into account __continuous output variables__ and give their estimates based on the difference between the __actual output__ and __predicted output__. Below are some metrics that are used for this prediction.

* [__Mean Squared Error__](https://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html)
* [__Mean Absolute Error__](https://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_absolute_error.html)

## Machine Learning Models

There are a large number of machine learning models used in the prediction of the demand for __Washington Bikes__. Below are the models that were used for prediction.

* [__Deep Neural Networks__](https://scikit-learn.org/stable/modules/generated/sklearn.neural_network.MLPRegressor.html)
* [__K Nearest Neighbors__](https://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsRegressor.html)
* [__Partial Least Squares (PLS) Regression__](https://scikit-learn.org/stable/modules/generated/sklearn.cross_decomposition.PLSRegression.html)
* [__Decision Tree Regressor__](https://scikit-learn.org/stable/modules/generated/sklearn.tree.DecisionTreeRegressor.html)
* [__Gradient Boosting Regressor__](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html)
* [__Linear Regression__](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LinearRegression.html)
* [__Long Short Term Memory (LSTM)__](https://www.tensorflow.org/api_docs/python/tf/keras/layers/LSTM)

## Visualizations

We would be mostly focusing on exploratory data analysis (EDA) to help us get an understanding of the data used. Based on this information, steps could be taken to perform feature engineering and improve model performance. 
 
<img src = "https://github.com/suhasmaddali/Images/blob/main/Input%20Data.png"/>

After reading the dataset with pandas, we get the following results about the data and the features.

Getting an understanding of the description of features in the dataset and understanding their spread in terms of variance and standard deviation.

It could be seen from the data that there is more data from the fall season category as compared to other seasons. But there is a negligible difference based on the count of various seasons.

Based on the results, it is shown that there is a higher demand for bikes during the fall season as compared to the summer, winter, and spring seasons. Therefore, giving this feature to our ML model would help determine the total demand for bikes based on the season.

Based on the average demand for bikes in various months, September had the highest demand followed by June and August. Hence, having information about the month can help determine the demand for bikes.

The recorded data is on a large number of days where there is a lower windspeed. This makes sense as cyclists tend to not ride bikes when there is a high wind speed.

The temperature distribution highlights that most of the records or bookings are done when there is a moderate temperature. Therefore, our ML model should be able to predict the demand for bikes well during moderate-temperature days.

This plot showcases the total demand for bikes during different hours of the day. In general, we tend to see a lot of bookings in the evening at 5:00 PM. This generally indicates that people go home after work from offices and they tend to book bikes during these times. In contrast, we tend to find less demand for bikes during early morning hours for timings from 12:00 AM to 6:00 AM.

### Model Performance

We will now focus our attention on the performance of __various models__ on the test data. Scatterplots can help us determine how much of a spread our predictions are from the actual values. Let us go over the performance of many ML models used in our problem of bike demand prediction.

[__Deep Neural Networks__](https://scikit-learn.org/stable/modules/generated/sklearn.neural_network.MLPRegressor.html) - After plotting the distribution of predictions on the test set versus the actual test output values, there is a lot of overlap in the model predictions with the true labels. Hence, deep neural networks are performing well on the test data. Let us also test the results of other models to determine the best model to be deployed to predict bike demand in the future.

[__K Nearest Neighbors__](https://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsRegressor.html) - The plot below shows the performance of K Nearest Neighbors on the test set. There is a lot of scatter between the predictions between the model predictions and the test data. Therefore, we can look for alternate models that can improve performance further.

[__Decision Tree Regressor__](https://scikit-learn.org/stable/modules/generated/sklearn.tree.DecisionTreeRegressor.html) - This model also performs well and its performance is comparable to that of deep neural networks. Depending on the device (Server, Mobile, and so on) that we use to deploy, we can switch to various models based on their performance and latency.

[__Gradient Boosting Regressor__](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html) - Gradient boosted decision trees (GBDT) also does a good job in identifying and predicting the demand for bikes during different times of the day. The performance is quite comparable to that of the decision tree regressor as shown in the figure below.

[__Linear Regression__](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LinearRegression.html) - Below is the performance plot of a linear regression model. Since the model is less complex, there is a higher chance that there could be bias in the model. This could be seen as there is not a straight line between the predictions and the true labels.

## Machine Learning Predictions and Analysis

* It is important to know some of the features that are present in the data so that we could be doing the __machine learning__ analysis.
* We performed various __data visualizations__ to understand some of the underlying features and once we got a good understanding of them, we used different machine learning models for predicting the demand for bikes based on these features.
* Once we get the machine learning predictions, we are going to be using different strategies that could aid us in the process of running the models in production that could be used in different ways in companies.
* Therefore, this would save a lot of time and money for the bike rental companies where the demand for the bikes is predicted beforehand with the aid of machine learning and deep learning respectively.

## Outcomes

* The models that were able to perform the best for predicting the bike demand are __Gradient Boosting Decision Regressor__ and __Deep Neural Networks__.
* __Exploratory Data Analysis (EDA)__ was performed to ensure that there is a good understanding of different features and their contribution to the output variable respectively. 
* The best machine learning models were able to generate a __mean absolute error (MAE)__ of about __23__ which is really good considering the scale of the problem at hand.

## Future Scope

* Additional features such as the __street connectivity score__ and __people's perceptions__ of the bicycling environment could be added to generate even more good predictions from the models.
* The best machine learning models could be __deployed__ in real-time where the demand for bikes is highlighted so that admins can take action based on the __demand__.

## 👉 Directions to download the repository and run the notebook

This is for the Washington Bike Demand Prediction repository. But the same steps could be followed for this repository.

1. You'll have to download and install Git which could be used for cloning the repositories that are present. The link to download Git is https://git-scm.com/downloads.
 
&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(14).png" width = "600"/>
 
2. Once "Git" is downloaded and installed, you'll have to right-click on the location where you would like to download this repository. I would like to store it in the "Git Folder" location.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(15).png" width = "600" />

3. If you have successfully installed Git, you'll get an option called "Gitbash Here" when you right-click on a particular location.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(16).png" width = "600" />

4. Once the Gitbash terminal opens, you'll need to write "Git clone" and then paste the link to the repository.
 
&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(18).png" width = "600" />

5. The link to the repository can be found when you click on "Code" (Green button) and then, there would be an HTML link just below. Therefore, the command to download a particular repository should be "Git clone HTML" where the HTML is replaced by the link to this repository.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(17).png" width = "600" />

6. After successfully downloading the repository, there should be a folder with the name of the repository as can be seen below.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(19).png" width = "600" />

7. Once the repository is downloaded, go to the start button and search for "Anaconda Prompt" if you have anaconda installed.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(20).png" width = "600" />

8. Later, open the Jupyter notebook by writing "Jupyter notebook" in the Anaconda prompt.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(21).png" width = "600" />

9. Now the following would open with a list of directories.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(22).png" width = "600" />

10. Search for the location where you have downloaded the repository. Be sure to open that folder.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(12).png" width = "600" />

11. You might now run the .ipynb files present in the repository to open the notebook and the python code present in it.

&emsp;&emsp; <img src = "https://github.com/suhasmaddali/Images/blob/main/Screenshot%20(13).png" width = "600" />

That's it, you should be able to read the code now. Thanks.

Washington-Bike-Demand-Prediction

Washington-Bike-Demand-Prediction 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Washington-Bike-Demand-Prediction

Washington-Bike-Demand-Prediction 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README