Predicting Nairobi Traffic Jams using Machine Learning

This is an interesting use of neural networks. Some roads are quiet long e.g Ngong Road, it would be beneficial to train the neural network on different sections of the road — that would be more beneficial to users.

A good approach would be to divide major roads in between major intersection (which normally cause most traffic), give a standard naming e.g JA (Jogoo A), JB(Jogoo B), JC (Jogoo C) et cetera, and use them as the target variables in the supervised learning using neural network.

A possible problem that you may run into is the curse of dimensionality — this occurs when you have many target variable and little data for each of the target variables. A remedy would be to use other ensemble learning algorithms such as Random Forest.

Google breaksdown it’s traffic prediction for road sections which I find useful especially if one is driving. An interesting approach would be including a rerouting algorithm that finds the optimal route given traffic conditions.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.