Each sector of this page takes to Kaggle. notebook where you can find the codes along with explanations and analysis.

Breast Cancer Detection using Logistic Regression

I used logistic regression to determine if the diagnosis is M or B (malignant or benign) by analysing other physical attributions of the person. Visit Kaggle.

Code and Resources Used	Heat Map	Results
Language: Python Packages used: numpy, pandas, matplotlib, seaborn, and sklearn Category: Logistic regression		The acuracy of the model is 96%, estimated from ROC curve.

Analysis of Donald Trump Speech in Dallas using Text Mining and Text Analysis

I used text analysis to create a word cloud that illustrates the main topics he mentioned in his speech. Visit Kaggle.

Code and Resources Used	Word Cloud	Results
Language: R Packages used: dplyr, tidytext, stringr, wordcloud, knitr, DT, tidyr Category: Text mining and analysis		I created a word cloud and datatable for the speech.

Happiness distribution analysis using Association Rule Technique

I used an apriori model predict find the assosciation among services in a place to predict the likelyhood of considered a place happy. The model is useful for residents to make a prediction for better housing environment. Visit Kaggle.

Code and Resources Used	Plot of the Rules	Results
Language: R Packages used: arules, arulesViz Category: Assosciation		I built 2 models with a support of 0.05 and 0.04 and a confidence level of 0.85 and 0.9 respectively.

Heart Disease detection using Random Forest Technique

I used random forest to determine whether the person has a high risk of cardiovascular disease or not. The model is useful for doctors to make a prediction and recommend the user on the best option for testing. Visit Kaggle.

Code and Resources Used	Plot of the model	Results
Language: R Packages used: caTools, FSelector, randomForest Category: Classification		The model trained was 78% accurate. The accuracy was calculated from the confusion matrix.

Diabetes Detection using Decision Tree Technique

I used decision tree to determine whether the person has diabetes or not. The model is useful for doctors to make a prediction and recommend the user on the best option for testing. Visit Kaggle.

Code and Resources Used	Plot of the model	Results
Language: R Packages used: caTools, FSelector, party Category: Classification		The model trained was almost 78.5% accurate. The accuracy was calculated from the confusion matrix.

Weather Forecast and Air Quality using Linear Regression Technique

I used linear regression to determine the temperature of a day given. The model is useful for predictions assist in weather forecast. Visit Kaggle.

Code and Resources Used	Plot of errors of the model	Results
Language: R Packages used: Amelia, caTools, corrplot, Metrics Category: Regression		The model trained has an error of 5.6. RMSE (Root Mean Squared Error) error was used for this model.