The Institute was extremely proud to host Pradeepta Mishra, Chief Data Scientist at Lymbyc, where he presented a Webinar on “Data Driven Product Development”. He has authored two books; ‘R data mining blueprints’ and ‘Mining text, sentiment, web using R’. He has also delivered a TEDx talk on ‘Can Machines Think?’, a session on the power of AI in transforming industries and changing job roles across industries.
Here are the 5 things you should know about Data Driven Product Development –
- Data Driven is not the same as Products that use data. It’s Data Driven when you make changes in your decisions based on the data. More importantly, these changes should help drive the predicted behaviour from the user.
- It is very important to find a “Problem” for a “Use Case”. Once, you have the problem, it needs to be translated mathematically, so that a solution can be found and converted into a business solution.
- Iterate the Product Development Cycle till you achieve perfection. Run parallel experimentation if needed to find the best solution with the help of an algorithm.
- The more the number of algorithmic layers, the slower the application will be. It’s important to make sure that a balance is maintained when it comes to the user experience.
- Deep Learning v/s Machine Learning – Go for Deep Learning when you need accurate results for complex problems, but when you need to explain how the model works, go for Machine Learning