Mithilfe von Machine Learning und Open Data zu Unfällen in Berlin (2018-2021) beantworten wir folgende Frage: Was sind die wichtigen Faktoren/Einflüsse auf Unfallgefahr? Und wie gut lässt sich damit die Unfallschwere überhaupt vorhersagen?

This research processed a fake news dataset, using TF-IDF and Count Vectorizer for feature extraction and evaluating multiple ML models through stratified cross-validation. Logistic Regression with TF-IDF was selected as one of the best models and further explained using LIME for interpretability.

Applying Convolutional Neural Networks (CNN) for recognizing manuscript digits.

Malicious URL detector built with deep exploration on feature engineering.

Read in a mnist file and using a neural network, predict the classifcation of 0-9 throughout the mninst dataset.

Deep Learning vs Tranditional ML methods for TB Drug Resistance prediction from Genomic data

Data sampling library

Churn prediction means detecting which customers are likely to leave a service or to cancel a subscription to a service.

Human Activity Recognition (HAR) has a wide range of applications due to the widespread usage of acquisition devices such as smartphones and its ability to capture human activity data.

This project consists in using machine learning to analyze the factors that affect wine quality and in building a model for predicting it. The model was tested on unseen wines to evaluate its accuracy.

[Modeling] Project in 2022 - Simple Model of important factors in the incidence of heart disease and prediction model

This python program demonstrates image classification with stratified k-fold cross validation technique.

Melanoma is a deadly disease. Image analysis tools that automate the diagnosis of melanoma will improve dermatologists' diagnostic accuracy.

Machine Learning App in R for predicting whether a newly released movie will be a hit or a flop. Practically useful for streaming services and cinemas.

使用比赛方提供的脱敏数据，进行客户信贷流失预测。

Performed univariate and bivariate analysis to understand the features and their relationships for loan approval prediction. Achieved highest accuracy of 98% for Extreme Gradient Boosting among all tested machine learning classification models.

This project aims to understand and implement all the cross validation techniques used in Machine Learning.