Predicting Jakarta Travel Time with 372 Features and One Linear Model
A single OLS model turns Jakarta trip context into 372 features, removes 157 extreme residual cases, and produces reproducible travel-time predictions.
I'm Mukhtada, an Information Systems student at the University of Jambi. I build full-stack web products, AI tools, data projects, and published research.
Opening screens from web, research, community, hackathon, and thesis projects.
Three projects from Mukhtada's research, AI tooling, and community work, with source links and technology details.
A local-first natural-language assistant for search, summaries, and communication, with an emphasis on direct, natural responses.
A Microsoft Word add-in for AI-assisted drafting inside the document, connecting Office.js with the writing workflow.
An e-ticketing system for Kerinci Seblat National Park, built for faculty research to replace on-site reservations. Contributed across front end and back end.
Writing on web development, AI tooling, open-source software, research, and the decisions behind real projects.
Writing on web development, AI tooling, open-source software, research, and the decisions behind real projects.
A single OLS model turns Jakarta trip context into 372 features, removes 157 extreme residual cases, and produces reproducible travel-time predictions.
Perbandingan harga BBM Pertamina DKI, biaya isi penuh, rupiah per kilometer, 50 km, dan 100 km untuk lima motor dari BeAT hingga ZX-25R.
Google confirmed a global spam update on August 18–21 but did not name AI content as a target. The policy, volatility data, and diagnostic limits point to a narrower answer.