Abhishek Purohit's repositories
Battle-of-Neighborhoods
In this project, I've tried to address the Business Problem of Restaurant owners in Ontario, Canada, who want to expand their business by opening up a new restaurant in Toronto. Based on my analysis, I've recommended the neighborhoods in Toronto where they'll earn maximum profit and face less competition. I've used FourSquare API, BeautifulSoup, for collecting the data. Using Python code, I tried to analyze that data using K-means Clustering Algorithm, and the final result was narrowed down from 39 to 4 neighborhoods.
Breast-Cancer-Prediction
In this project, I have built a classifier to predict the stage of Breast Cancer in a patient, whether it is Malignant or Benign. I've applied different classification algorithms on the data, including KNN, Logistic Regression and Random Forest, and obtained the classifier giving the best accuracy among them.
Titanic-Machine-Learning-from-Disaster
In this project, I have built a classifier to predict the survival of a person on the historical Titanic ship. I've loaded the data from Kaggle, cleaned the data, and applied different classification algorithm on the data. Obtained the Best classifier using Stratified Cross Validation.
Computational-Method-of-Analyzing-Building-Structures
Designed a model of a N-storeyed moment resisting reinforced concrete frame building in SAP2000. Analyzed it for different load combinations as per IS 456 and verified the results using a MATLAB code
The-Best-Classifier-Machine-Learning-Project
In this project, I have built a classifier to predict whether a loan case will be paid off or not. I've loaded a historical dataset from previous loan applications, clean the data, and applied different classification algorithm on the data. Following algorithms are used to build the models: k-Nearest Neighbour, Decision Tree, Support Vector Machine, Logistic Regression. The result is reported as the accuracy of each classifier, using the following metrics when these are applicable: Jaccard index, F1-score and LogLoss.