yukaberry/portfolio_forecasting_diabetes

To diagnostically predict whether or not a patient has diabetes, based on certain diagnostic measurements

Project : Data Science

Forecasting Diabetes by diagnostic measures

To diagnostically predict whether or not a patient has diabetes, based on certain diagnostic measurements. All patients here are females at least 21 years old of Pima Indian heritage. Here to see the project findings

1. The objecttive of this project
2. Data details
3. Model and methods Overview
4. Domain (Diabetes types)
5. Feature creattion and Data cleaning
6. Model selection and Evaluation
7. Challenges and Augmentations

Motivation

Learning basic techniques and approach to solve a classification problem by machine learning.

Files' and folders' descriptions

• data A data folder containing raw and processed data files

• images Images kept in this folder for the project findings

• project_findings Report of this project

• utils

  • cleaning_utils.py
  • modeling_utils.py

• notebooks clearning.py and turning_model.py's notebook version. It shows graph and visualisation process.

• clearning.py Preprocessing datasets

• turning_model.py Evaluating and turning a model

• turning_resampled_modeling.py Evaluating and turning a model which was trained by resampled dataset

• requirements.txt Software Requirements