Valor is a centralized evaluation store which makes it easy to measure, explore, and rank model performance.

Summary Evaluation Tool

The purpose of this project is to develop and compare two machine learning models to detect spam emails. Spam detection is a crucial task in email filtering systems to protect users from unwanted and potentially harmful emails. The project involves using a dataset containing various features extracted from email content.

Python tools for the AeroCom project

Microsoft Stock Price (closing) Prediction using Stacked LSTM and ARIMA (6,1,6) models

"SocialScope harnesses the power of data science to Instagram's vast content, providing insightful analytics and trend predictions for informed decision-making."

an MLOps/LLMOps platform

Tool to simplify the collection of datasets from openml.org. Includes support for timeouts, filtering and local caching of files, and for running tests against multiple datasets. Allows both random and reproducible tests. Supports reading from cache when openml.org or the internet is unavailable. Multiprocessing is supported for faster compariso…

Machine learning research project predicting and identifying cyberattacks on the SWaT testbed from iTrust Centre Labs.

A lambda to invoke inference requests to AWS SageMaker endpoints, using test data previous created. Helping with monitoring model performance.

A project designing and evaluating the fairness of a predictive model for arrest decisions using the North Carolina Policing Dataset. The study compares logistic regression, KNN, and fine-tuned KNN to ensure high accuracy and fairness in predictions based on gender, age, and race.

Repositori ini berisi dua project analisis data menggunakan metodologi CRISP-DM. Project pertama meneliti tren penjualan Walmart dengan Supervised analysis menggunakan algoritma Naive Bayes Gaussian dan K-Nearest Neighbors. Project kedua mengeksplorasi faktor sosio-ekonomi antar negara dengan Unsupervised analysis.

Tools created for machine learning classification model evaluation

Content: Peformed EDA on Decathlon 2009-2011 dataset, preprocessed & cleaned the data, analysed critical KPIs & forecasted sales for the coming year using Triple exponential (Holt_winter) model.

Prediction of students' dropout using classification models. Data visualisation, feature selection, dimensionality reduction, model selection and interpretation, parameters tuning.

This repository contains a machine learning model aimed at predicting student performance across various metrics. Utilizing a diverse set of Machine Learning Regression algorithms, the model predicts scores based on demographic and academic variables.This project demonstrates robust approach to leveraging machine learning for educational outcomes.

Bias detection and contextual evaluation tool for your AI projects