Predicting/identifying segments 4 delivery time delays

Inspiration

[Vijay Rajan's] The Madurai Idli shop near my house is flooded with Swiggy Delivery boys begging for their =breakfast orders to be given to them for pickup on a Sunday morning. This got me thinking that there is room for prediction, forecasting, exploratory data analysis tools and monitoring at Swiggy. [Manohar Battula's] Being a fresher, I am diving headlong into Machine Learning. I would like to learn a lot and these hackathons push me to the hilt. Proud to mention that I came 15th among 100s of other contestants in Analytics Vidhya's hackathon last month.

What it does

We have the ability to discover patterns where the likelihood of delays in delivery are larger than usual. Did you know that when a discount is offered by a restaurant, the chances of delay in delivery are higher than global averages?

Using Predictive Analysis and a proprietary Exploratory analysis, we help you know where you fall short.

Of the 5029 orders delivered by Swiggy satisfying the following condition day_type=WEEK_DAY and discount_flag=1 and distinct_item_count=5_or_more_distinct_items_ordered and metropolitan_area=4nx, 1778 orders that is 35.35% of the orders were delivered in more than 50 minutes.

How I built it

SQL, A Lot of SQL, Java for text processing, Shell Scripting, R-programming and a solution in Java that is proprietary called GIST(Generalized Item Sets)