Predicting Traffic Density in Istanbul Using Machine Learning

Local Chapter Istanbul, Turkey Chapter

Coordinated by Turkey ,

Status: Completed

Project Duration: 30 May 2023 - 17 Jul 2023

Open Source resources available from this project

Project background.

Istanbul, the largest city in Türkiye, is infamous for its heavy traffic. The traffic density problem in Istanbul is a complex issue that stems from several factors such as population growth, urban sprawl, inadequate public transportation, and a high number of private vehicles on the roads. With over 15 million residents and growing, the city’s roads are often gridlocked, particularly during peak hours.

A study conducted by Yeditepe University revealed that in 2017, roughly 55% of the travel time for both drivers and passengers was wasted due to heavy traffic. What should have been a 20-minute journey in lighter traffic often stretched to approximately 45 minutes. Furthermore, during periods of heavy traffic, the average driving speed was found to be a mere 36 kilometers per hour.
Some of the data from the study are as follows:

– Istanbulites are sacrificing an average of 3.5 years of their lives for the delays caused by the intensity of the city traffic.
– Drivers moving on the arterial roads of Istanbul in 2017 moved at an average speed of 37 km / h during the day, and 54% of their time was lost due to traffic congestion.
– In the same arterial roads, the average speed of the drivers was measured at 26 km / h during the weekday mornings. 67% of the time the drivers spent on the road traveling in the morning was caused by traffic intensity.
– These values ​​increased even more during weekday evenings. Drivers moved at an average speed of 22 km / h and lost 71% of their time due to traffic congestion.

Project plan.

  • Week 1

    – Introduction
    – Setting a project outline
    – Data Collection

  • Week 2

    – Data Preprocessing and Exploratory Data Analysis
    – Feature Extraction

  • Week 3

    Model Development and Training

  • Week 4

    Model Evaluation

  • Week 5

    Final Documentation

Learning outcomes.

Time Series Forecasting, Data Extraction, API, Predictive Analysis, Data Visualization

Share project on: