Himanshu Gajanan Tikle

To associate with an organization that gives me a chance to prove my knowledge and efficiency and enhance my skills in the field of the technology and be a part of the team that excels in work towards the growth of the organization. Willingness to learn and adaptability to put learning into practice. Adaptable to environmental changes and emerging trends. Comprehensive problem solving abilities.

1.Indian Liver Patient Prediction: This Project consist of Liver Patient identification Data Set. 

Objective: The target is to identify whether the patient has an anomaly or abnormal functioning of the liver.

Summary: The Liver Pateint Prediction Dataset contains liver patient records and  non-liver patient records collected from North East of Andhra Pradesh, India. After observing the dataset we know that it is an imbalance dataset so to overcome this I applied the Smote as to Balance dataset by doing Over Sampling. After  EDA come to know it is a Classification Type of Dataset after which Applied many Machine learning Algorithms as Logistic Regression, Naive Bayes,K-Nearest Neighbors,Support Vector Machine, Random forest and Descision Tree. Out of which we get the K-Nearest Neighbor as the best fit Algorithm for the Dataset giving 85% Accuracy.

2.Game AI - Winner Prediction: This is a PUBG game Data Set. The dataset contains a large number of anonymized game stats for a single player with all match types containing of 10 Lakh Players Data.

Objective: The target is to create a ML model which predicts players' finishing placement based on their final stats.

Summary:  The Game AI Prediction Pubg Dataset containing all kind of Matches (Single Player,Groups etc). After Analyizing and doing EDA got to konw that Data is a Regression Type and Applied Different Machine Learning Algorithms. Out of them the Linear Regression and  XGB Regressor were Performance was good .Model Scores out of which we got  93.01% as the highest accuracy score from  XGB Regressor. So we conclude that  XGB Regressor is the best fit model for Predicting the Game Winner Place.