Research/Blog
Meeting Minutes from AI Lab Hands-On Workshop on Saturday 14th Sep in Bengaluru
- September 19, 2019
- Posted by: vsinghal
- Category: Machine Learning
#CellStratAILab #disrupt4.0 #WeCreateAISuperstars
Our AI Research Team Lead Indrajit Singh presented a fantastic workshop on Retail Demand Forecasting using Time-Series Modelling.
The dataset was the Food Delivery dataset provided by the Genpact Hackathon challenge. Here is the summary of the solution :
Machine Learning is not new and so is the time series. In the real world scenario, solving problems like demand forecasting needs a right intiution and knowledge to execute models in production. Keeping this in mind, this guided session on Food Demand Forecasting was designed to help the community get a better understanding of the execution steps and challenges in demand forecasting.
A demand forecast model is not something one can build blindly. Before diving into the algorithms, the data scientist should first determine what it will be used for (e.g., buying optimization, store allocation, pricing optimization, etc.).
The dataset comprised of weekly demand data for various food delivery centres, city and region information for these fulfillment centres, and meal categories. The model used a somewhat uncommon evaluation metric called the RMSLE or Root of Mean-Squared Logarithmic Error.
Once we clean the data and identify the features, one important aspect is to choose the right algorithm for our model. There are popular ML boosting algorithms like XGboost, ADaBoost & GBM, but we have used CatBoost here for solving the demand forecasting problem. Catboost is a relatively new regression and classification algorithm developed by the Russian firm Yandex. Catboost has the primary advantage of taking care of the categorical variables over other boosting algorithms. Apart from that, it’s easy to use, robust and very competitive with any leading machine learning algorithms.
The audience enjoyed the workshop thoroughly and understood the intricacies of feature engineering in demand forecasting as well as how to use CatBoost for this task.
Wish to learn advanced AI ML ? Interested in joining the AI revolution ? Then engage with India’s No 1 AI Lab – the CellStrat AI Lab. To learn more, visit our AI Lab meetup this Saturday (21st Sep) in Bengaluru or Gurugram :-
Bengaluru AI Lab (Saturday 21st Sep) :-
Register : https://www.meetup.com/Disrupt-4-0/events/264758741/
Topic : Dynamic Programming, AutoEncoders
Loc. : WeWork, Embassy Tech Village, ORR, BLR
Presenters : Vivek Singhal, Niraj Kale, Gouthmanan Asokan, Shubha M.
Gurugram AI Lab (Saturday 21st Sep) :-
Register : https://www.meetup.com/Disrupt-4-0/events/264743997/
Topic : Dockers in ML, Image Manipulation Detection
Loc. : WeWork Forum, DLF Cybercity, Phase III, Gurugram
Presenters : Vishal Singhal, Sonal Kukreja, Akarsh Verma
See you this weekend for the AI Lab meetup! Lets disrupt the world with AI, together !
Questions ? Call me at +91-9742800566 !
Best Regards,
Vivek Singhal
Co-Founder & Chief Data Scientist, CellStrat
+91-9742800566