RUSH'19

Category : Web

1. Create a signup page where a user can register and a login page to log in with the details provided during registration.
2. Successful login should redirect to a profile page which should contain additional details such as dob, age, contact, etc that can be updated by the user.

Must follow below constraints:

1. Html, CSS, JavaScript, PHP should be in separate files.
2. CSS should be written using sass/scss. The project should contain both scss file and compiled css file.
3. The form submission should be handled using ajax, do not use HTML form action.
4. Use bootstrap to make page responsive.
5. Don't use more than two fonts and any icons or images should be in SVG format.
6. Use jQuery to simplify dom manipulations.
7. User data should be stored in MySql DB and also in the JSON file.
8. Use prepared statements while performing MySql queries.

Category : Machine Learning and AI

This challenge is aimed at analysing different parameters of a Food and Beverage company, to understand the interaction between various products, brands, outlets, commissions and their relationship to sales, revenue and market share.
You may choose to use any open source analysis tool(s), and any statistical or machine learning algorithm(s) that you deem necessary for your analysis.
The intent of this exercise is to get a firm understanding of the contribution of commission payouts to overall sales and market share of each product, brand and company through the use of effective exploratory data analysis, data visualization, predictive modeling and inference.

Constraints:

1) Solution should be data driven.
2) Should effectively use visualizations to present findings.
3) Should use statistics or machine learning models to firmly understand patterns and relationships in the data.
4) Document should explain the framework, design, implementation, inference and approach to each step.

Category : Machine Learning and AI

Build a system that identifies the type of question being asked by the user. Any given question can belong to four types: Who, What, When or Affirmation (Yes / No)

Any sentence that doesn't come under this category must be labeled as Unknown.

For testing your model, you need to build a dashboard that accepts questions, and classifies the same. Once that is done, build an interface to display it to the user, and validate your model's response with that of the user's actual intent.

Bonus: If you can improve the accuracy of your model based on the input given by the user.

Furthermore, the dashboard must show graphs with regards to the query being processed, the number of predictions vs the number of correct predictions. Your application will be evaluated based on the number of analytics that you can provide with the data you've gathered.

Dataset for testing the service will be provided on the day of competition.