Beast code in Giters

GeekerLee's starred repositories

Dow-Jones

Language:R100

The-Airbnb-Classification-Project

This project is from the Airbnb Recruitment Challenge on Kaggle. The challenge is to solve a multi-class classification problem of predicting new users first booking destination.

Language:R1100

Prediction-of-Life-Expectancy-using-WDI-Indicators

Trained Ridge, Lasso and Linear Regression models along with Decision Trees

Language:R100

Beijing-House-Price-Prediction

Desgined and Implemented Machine Learning algorithm to determine Important factors that affects pricing and predict prices

Language:R300

Data-Mining-Default-of-Credit-Card-Clients-

For this project, we have used a dataset published by UCI Machine Learning. The intention of choosing this dataset is to improve our knowledge on data mining concepts using real time data, building different models to predict the outcome variable and to interpret our result. The project mainly includes classification of 30,000 users into defaulters and non-defaulters. Classified using Logistic Regression (Logit), Neural Network and Classification Tree (CART). Able to get accuracy of 85%.

Language:R200

Prediction-of-rate-of-spread-of-Covid--19

• Performed feature selection using subset selection and shrinkage methods like Lasso and Ridge regression to identify the most contributing features impacting the spread rate of Covid-19 from the 66 predictors defining mobility changes in places of activity • Trained multiple classification models using Logistic regression, ridge regression, random forests and gradient boosting; Achieved a test mean square error of 0.00373547855 by hyperparameter tuning of the Xtreme gradient boosting model.

Language:R100

Analysis-of-Boston-Housing-Data

Language:R300

Machine-Learning-in-Econometrics

Collection of lecture notes and excercises for a course "Machine Learning in Econometrics"

Language:HTML1800

Major-Project

Machine Learning is a vast field and has provided solutions to many. It even lays feet in the world of stock market as many brokers use Machine Learning algorithms to predict the value of the stock client wants to invest in. The project tends to show which of the algorithms or their hybrids are best for stock predictions. The project covers algorithms such as Linear Regression, Decision Trees, Random Forest Regression, Artificial Neural Networks and Recurrent Neural Networks. There are various research papers for stock predictions using Artificial Neural Networks and Recurrent Neural Networks, but the classical algorithms are not used, so it will reflect the results of those algorithms as well. The purpose of the project is to analyze the machine learning algorithms on the large dataset of different stock prices and their categories in training set part of dataset to predict the prices of stock, so that we can decrease human effort in deciding the prices of stock and come up with better guidance in order to help people to decide which stock to invest in so as to maximize the return and also to know which algorithm or hybrid of various algorithms gives a better prediction result.

Language:R100

Machine-Learning-in-Stock-Price-Prediction

The file “mf850-finalproject-data.csv” contains monthly stock return data for publicly tradedcompanies in the United States during the years 2015 through 2017. The file also containsfundamentals data characterizing each company in each month, the monthly returns of the S&P500 index, and some monthly sentiment measures. Your first task consists of constructing forecastsof the monthly stock return of a company based on the characteristics of the company and themarket. Your second task consists of constructing predictors of whether a stock will grow or fallover the course of a month based on the characteristics of the company and the market.

Language:R100