Ineuron-Internship

Mushroom Classification Project

Problem Statement:

Mushroom is one of the fungi types foods that has the most potent nutrients on the planet. Mushrooms have major medicinal advantages such as killing cancer cells. This study aims to find the most appropriate technique for mushroom classification, and mushroom will be classified into two categories, Poisonous and Edible.

Where the dataset contains different features of the mushrooms like gillcolor, sporeprintcolor, population, gillsize, stalk root, bruises, stalkshape etc. The proposed approach will implement a different techniques and algorithms like Decision Tree, Random Forest and Boosting Techniques. Random Forest performed well with 100% accuracy. After we have done hyperparameter tuning and it is deployed on Amazon Web Services (AWS) platform.

Approach used

The main goal of this project is to predict whether the Mushroom is Edible or Poisonous

 
 Data Exploration     : I started exploring dataset using pandas,numpy,matplotlib and seaborn. 
 Data visualization   : Ploted graphs to get insights about dependend and independed variables. 
 Feature Engineering  :  Removed missing values and created new features as per insights.
 Model Selection I    :1. Tested all base models to check the base accuracy.
                       2. Also ploted residual plot to check whether a model is a good fit or not.
 Model Selection II   :  Performed Hyperparameter tuning using gridsearchCV and randomizedSearchCV.
 Pickle File          :  Selected model as per best accuracy and created pickle file using joblib .

## Technologies Used

1. Python
2. Sklearn
3. Flask
4. Html
5. Css
6. Pandas, Numpy
7. Database

Deployment

link : http://classification-project-api.azurewebsites.net/