How to build a Language Detection model using Multinomial Naive Bayes algorithm in Python.

how to perform anova test in rstudio

This is a Synthetic Financial dataset which is a simulator generated dataset of mobile money transactions generated for fraud detection research. The dataset has more than 10 lakh observations with 11 features. EDA was performed on the dataset to draw insights and interpret the business. Then built two models one with using Random Forest algorithm and the other with Extreme Gradient Boosting algorithm to cross-check the accuracy and sensitivity of the models.

Building a multinomial logistic regression on principle components

Classification with Naive Bayes algorithm

ARIMA Time Series model to predict future flow of air travelers.

Building predictive models both for classification and regression using KNN algorithm.

Predicting the outcome variable only with significant variables.

This model is to find out whether overfitting is present in the model by comparing train and test misclassification errors.

The importance of Robust and Resistant regressions in dealing with influential observations and outliers.

To look at how these three methods impose penalty on the size of the coefficients to reduce the variance of the estimates.

Comparing simpler model with complex model in order to use the best model.

Finding out how mother's IQ is playing a significant role in child's IQ.

Predicting miles per gallon of cars using weight variable as well as finding Whether weight of a car is playing a significant role in mileage of a car.

How to form clusters using Hierarchical and Non-Hierarchical methods

Binary data with 4 variables - Admit, GRE, GPA, and Rank. GRE and GPA scores and Rank of school where a student has come from. Using these independent features need to predict whether a student’s application was rejected or accepted.

How to choose the best SVM model

Classification model using Random Forest algorithm. The outcome variable NSP has three classes where class 1 belongs to normal condition, class 2 belongs to suspect condition, and class 3 belongs to pathologic condition of a patient. Further, the target variable, NSP, is predicted using 21 independent variables.

Classification