/Time-Series-forecasting-for-the-number-of-taxi-passengers-in-manhattan

To assist the operation team of the NYC Taxi commission in Manhattan to optimize the distribution of the taxi fleet

Primary LanguageJupyter Notebook

Time-Series-forecasting-for-the-number-of-taxi-passengers-in-manhattan

To assist the operation team of the NYC Taxi commission in Manhattan to optimize the distribution of the taxi fleet

How to Run the code :

  1. Download the code (Contact me if you want the token to download via linkedin)
  2. Install the requirement with this syntax : pip install -r requirement.txt
  3. Run the ipynb file in jupyter notebook

Analyze

Business Problems : Operation team of the NYC taxi commission want to optimize the distribution of the taxi fleet in 1 month ahead to maximize the revenue.

To solve that problem, I make a time series forecasting model using prophet library. now let's see the dataset

image

Dataset have 19 columns with 4 columns have null value, but the columns that have null data is not important in our analysis, so we dont need to do anything from that columns.

Now let's make a timeseries model

  1. Prepare the data so that it is ready to be used to create a mode
  • Take a important columns that you need, in this case we need pickup datetime and location and order quantity. (in my code the data was in variable called final)

image

  1. Choose one location that you want to analyze (in my case I Choose location with ID 4)
  2. Make a time series model

image

  1. Make a prediction

image

  1. Last, Analyze data from prediction

image

From the data above, we can recommend to Operation team of the NYC taxi commission to send about 15-20 taxis between 23:00 to 03:00 everyday and besides these hours it is enough to prepare 10-12 taxis