-- This is a FREE event--
Bangalore AI/ML Meetup invites you to this special meetup to get you started with Artificial Neural Network and Deep Learning.
10:15 AM - 10:30 AM: Registration & Welcome / Introduction
10:30 AM – 12:00 Noon: " Hands-on cum Introduction to Artificial Neural Network and Deep Learning" by Pramod R (Senior Lead Data Scientist at Fidelity Investments) (https://www.linkedin.com/in/pramod-r-05b38212/)
Description for Talk 1:
Inclusions for the first Hands-on on ANN and Deep Learning
● What is a Neural Network?
● Popular use cases/examples of Neural Network
● Deep Dive:
○ Discuss Linear Separability/non Linear Separability
○ How a perceptron compares to the human brain
○ Components of neural network - input, output and hidden
○ What is inside the hidden layer - weight+bias and the activation
○ What is a Loss function
○ Gradient descent and back propagation with example
○ Activation function
● Hands on Keras -
○ Build a feed forward network for a classification task
○ Measure the accuracy of the model
○ Save the model and tips on deployment
● What next from here - Preview into RNN and CNNs and upwards
- Laptop with internet connectivity & Google account
12:00 Noon – 12:15 PM: Tea Break
12:15 PM - 1:15 PM - "Creative techniques in the Deep-Learning world" by Samiran Roy (Data Scientist at Envestnet | Yodlee) (https://www.linkedin.com/in/samiranroy/)
Description for Talk 2:
A data scientist with variety of experience in the industry typically gets to learn about a) The basics of neural networks and b) Common architectures like CNN, RNN, LSTM, Auto-encoders.
It is easy to forget that Deep Learning is rapidly evolving, and people use neural networks in very creative ways to solve challenging problems.
Speaker would be showcasing some lesser known neural architectures that he has worked on as a Data Scientist at Envestnet | Yodlee. The architectures are applicable to a wide range of problems in the industry.
The talk will broadly include:
- Deep Learning review
- Reinventing the neural network paradigm
- Can we use a neural network to:
a) Learn a simple mapping: if the input is x, the output is also x
b) Sort numbers
c) Have a image of a white square -> Black out one pixel at random, try to predict the coordinates of the black pixel
- Semantic Hashing for Approximate Nearest Neighbour Search
- Siamese Networks for Learning Distance Measures
- Multi Modal Deep Learning for Joint Embedding Spaces
- Neural Networks that operate on Graphs
Note: Laptops are allowed at this venue. Please bring your laptop to do the hands-on exercises in parallel.
Please do not RSVP / Waitlist here. Please register in the konfhub page: https://konfhub.com/aimeetup
Heartiest thanks to Nutanix for hosting / sponsoring this meetup.