/ai-singapore-khalishah

Primary LanguageJupyter Notebook

This folder includes files for application to AI Singapore Apprenticeship Programme for Khalishah Nadhirah Bte Abu Bakar.

Files include

  1. mlp folder
  • Contains 2 files
    1. mlp_proj_working_file.ipynb - File where I worked on the model training and model tuning
    2. mlp_main.py - File which contains functions to run the machine learning pipeline
      • Function 1 - data_prep()
      • Function 2 - traffic_model(df)
      • Function 3 - main() (elaboration of functions below)
  1. README.md
  • Description of files
  1. eda.ipynb
  • This file contains the exploratory data analysis of traffic_data.csv.
  1. requirements.txt
  • This file contains the required Python dependencies for the files in mlp folder.
  1. run.sh
  • Executable file to run the data prep and models.

mlp_main.ipynb contains:-

  1. data_prep() function executes the following:
  • Import necessary packages
  • Read data from URL
  • Parse dates in dataframe and extract date features from date field
  • Fixing in accuracies in data like strings case, granularity of dataframe
  • Extract information from Holiday feature
  • Encoding of selected categorical features
  • Drop unnecessary features
  • Returns final dataframe to be fed into the model
  1. traffic_model(df) executes the following, taking cleaned dataframe as an input:
  • Import necessary packages
  • Get independent and dependent variables
  • Train Test Validation Split
  • Specify metrics for validation
  • Specify functions to print scores
  • Specify final model with final chosen training parameters
  1. main() executes both data_prep() and traffic_model(df)
  • Runs the whole pipeline