MACHINE LEARNING PIPELINE (PREDICT ACTIVE USERS FOR E-SCOOTER RENTAL)

This is a python project submission for AI Apprenticeship Programme Technical Assessment.

This project is for the prediction of active users ('guest-users' and 'registered-users') for an e-scooter rental service in a city.

The following features are provided:

Independent features: Independent Features:

• date​: Date in YYYY-MM-DD
• hr​: Hour (0 to 23)
• weather​: Description of the weather conditions for that hour
• temperature​: Average temperature for that hour (Fahrenheit)
• feels-like-temperature​: Average feeling temperature for that hour (Fahrenheit)
• relative-humidity:​ Average relative humidity for that hour. Measure of the amount of water in the air (%)
• windspeed​: Average speed of wind for that hour
• psi:​ Pollutant standard index. Measure of pollutants present in the air. (0 to 400)

Target Features:

• guest-users​: Number of guest users using the rental e-scooters in that hour
• registered-users​: Number of registered users using the rental e-scooters in that hour

data url: https://aisgaiap.blob.core.windows.net/aiap6-assessment-data/scooter_rental_data.csv

Table of Content

1. Overview of the machine learning pipeline
2. Running of the machine learning pipeline
3. Configure your own machine learning pipeline!

1) Overview of the machine learning pipeline

Step a) Data-preprocessing

After the data is imported, the data is preprocessed based on our findings from exploratory data analysis. (file: eda.ipynb)

Step b) Splitting into train and test set

The data is then split into training set X and test set y.

Step c) Encoding categorical features

After step 1, 'weather' is the only categorical feature remaining. Due to the nature of the feature (i.e. not ordinal and only has a few unique values), one hot encoder was used for the encoding process.

Step d) Normalization/Scaling of the data

MinMaxScaler was used as it preserves the shape of the dataset.

Step e) Training on Machine learning model

Multiple models were trained using GridSearchCV find the model that scored the best on "r2 - Coefficient of determination".

Supervised Learning Regression Models used: 'LinearRegression' - Standard OLS 'LassoRegression' - OLS with regularization (introduce penalty = absolute of the maginitude of the coefficient) 'RidgeRegression' - OLS with regularization (introduce penalty = square of the maginitude of the coefficient) 'XGBRegression' - gradient boosted decision tree (objective function with training loss and regularization)

Step f) Results

The machine learning pipelin will provide you with the following results

• Model performance table (on the training set) Table tabulating each model being trained, its performance based on scoring selected, and the best parameters that returned the scoring.

• Prediction report (on the test set) Adj R-squared and Variance between prediction results and test set

• First 30 predictions (on the test set)

2) Running of the machine learning pipeline

Machine Learning model created in with Python version 3.6.7/3.6.8 and bash script.

Installing Dependencies

Paste the following command on your bash terminal to download dependencies

pip install -r requirements.txt

Running the Machine Learning Pipeline

Paste the following command on your bash terminal to grant permission to execute the 'run.sh' file

chmod +x run.sh

Paste the following command on your bash terminal to run the machine learning programme

./run.sh

3) Configure your own machine learning pipeline!

A configuration file (file: ./mlp/config.py) was included to allow anyone to make their own configuration to the pipline. Users can make their own configurations to steps(b-e) mentioned above.