The Modeling For Predict The Vehicle Count In Rhode Island

Musthofa Syarifudin

19 | Indonesia | Mathematics


The goal of the project is to be used to reduce traffic jam in the smart city. project was develop in python and jupyter notebook with the keras and tensorflow library to help coding. to reach the goal the model need to be integrated with optimise traffic system and other guide system like google maps. the model could also develop to predict fuel emmission and fuel need by estimates the number of car and calculate average fuel need per car and average fuel emmission per cars. The project was meant not only to reduce traffic jam but also developing automated system where the car will look for least populated street for the ride. i really hope the model could be optimise for traffic system and apply in smart city. the project was using LSTM or long short term memory which using 3 input as factor and produce 2 output. the input consist of the new data, the last data, and the last weight which help the system to learn over the time. the output of the model consist of data and weight. LSTM has advantage over other type of machine learning model, they could predict better accuracy, and more efficient. the model also using dropout layer to prevent over fitting of the model, which turn down 20% of the neuron to start refresh in data. the data that enter the model will be single from timeseries format, using the model we calculate the prediction of car number in the next day. the model will be helping costumer in numerous ways . first the model could help people to take a bus rather than using car which will faced by traffic jam, second the costumer will benefit with integrated optimise system to reduce time in street with optimise traffic system, third the model could be the choice for with road is full capacity and which is not full.


Hello my name is musthofa you can call me thofa for shorts. i was beginner programmer and data science. i attend mathematics at sebelas maret university in the 2018. Since forst i was interested in developing machine learning model, android, and data science. especially when data could be using to increase our value in life, helping other, and solving problem.


Vehicle Counts / RI 5, Cranston, Rhode Island, United States / 2017