Reference no: EM133417477

Group Project - Developing a predictive machine learning model (classifier) and deploy it as a web API for inference

1) Purpose

The purpose of this project is to:

1. Pre-processing  - Retrieve  & prepare the data:
   1. Load and explore the dataset referenced in section 4 in this document using techniques learnt during this course.
   2. Visualize the data and describe it thoroughly, identify correlations..etc.
   3. Clean, transform categorical data and model the dataset using the techniques learnt throughout the course in preparation for building a predictive model.
2. Model building & fine tuning
   1. Build a supervised predictive model based using a suitable classification algorithm(s) in python , utilizing scikit-learn, pandas, numpy...etc. To provide predictions as specified in project specification, section 3 of this report.
   2. Validate / score and evaluate the models and choose the best model after carrying out hyper-parameter tuning.
3. Model deployment
   1. Build an API for the model using Python Flask framework.
   2. Deploy the model on local host.
   3. Build a simple front end to access the API and pass new feature values to the prediction model for inference.

2) Guidelines & Instructions

Be sure to read all the

General:

- This Project is to be completed in groups of 5 students.
- Read the textbook, course lecture content, class examples, and additional references provided here. Each team are free to research and use more materials and tools to implement a good solution, just make sure to reference it in your solution and your report.

- Presentation:

- Each group will have to present and demonstrate their solution in Week 14 or as agreed with their professor.
- During the presentation, each group needs to demonstrate the solution i.e. live code execution and as needed illustrate part of the code. In addition, the group needs to present all key findings, assumptions, constraints, and the list of technologies used clearly.