The first round of Machine Learning Developers Summit– (MLDS) 2020 concluded in Bengaluru, featuring more than 50 engaging sessions, keynotes and workshops from some of the leading minds in the field of machine learning and artificial intelligence. The conference covered both the technical and business aspects of machine learning, attracting developers as well as C-suite executives from all over India.
Insightful sessions helped attendees understand how AI/ML impacts the world, both in terms of career and business prospects. The workshops and research paper presentations garnered huge crowds who explored the latest in machine learning technology and tools hands-on. The brainstorming was incessant — from tech talks and workshops to paper presentations and casual conversations. Here are glimpses from some of the sessions at MLDS 2020 Bengaluru edition:
One of the most awaited hands-on sessions from MLDS was by Viral Shah, the co-creator of Julia programming. The session was based on the next big trends– AI 2.0 and Differentiable Programming, and how to implement machine learning using Julia programming language.
Then, we had the talk “More Data Science, Less Engineering with Netflix’s Metaflow” by Savin Goyal from Netflix. Savin demonstrated how to use Metaflow to make the ML process reproducible and work across domains. Metaflow is an open-source framework for data-science projects built by Netflix Research.
Two young developers from IBM Software Labs, Pragun Shukla and Rachit Chaudhary demonstrated their innovation called Virtual Aid as part of Global Call for Code– which aimed at solutions for natural disaster preparedness and relief. The two developers showed about how they built VirtualAid, a programmable drone with RaspberryPi Camera and Watson Visual Recognition.
Among one of the workshops at MLDS 2020, Mohanraj Vengadachalam, Data Scientist at Renault Nissan presented hands-on training on Convolutional Neural Networks (CNN) using TensorFlow and Keras on Google Colab Platform. His workshop also included the various CNN architectures and Transfer Learning Techniques for pre-trained CNN models.
Further at MLDS 2020, ZS Chief Technology Officer Abhishek Trigunait and Data Science Manager Sagar Madgi gave an interesting talk about productisation of ML models and automatic feature selection. The duo discussed ‘automated feature engineering’ as a viable alternative to handle the required complexity of predicting customer journey events and possible gaps.
Analyst- Data Science Nithya Vasudevan gave her paper presentation about emotional stress detection using deep learning by taking an in-depth look at the emotional and mental health of India as a country.
Data science leader Mathangi Sri presented an insightful session on Anthropomorphism in Conversational UI where she talked about the gap in human and machine communication and how the interaction can be made better using techniques that are based from human psychology.
Vinodhini Ranganathan, Senior Data Scientist at Cisco gave a well-received session on fake review detection, which started off with classic quote on data by Edward Deming. She highlighted the innovation in detecting the ever-increasing problem of fake reviews. She covered annotating fake reviews, a general framework for identifying fake reviews, and NLP techniques that can be leveraged for identifying fake reviews.
Sundara Ramalingam N, Head of Deep Learning practice at NVIDIA India spoke on how to create the successful recipe for building AI into the existing traditional workflows of an organization. He covered the recent advancements in AI technology from both Infrastructure and Software perspective and recommended best practices learnt from multiple domains for setting up a successful AI practice.
The attendees had one-on-one reach with prominent machine learning researchers and experts, and to ask specific questions on technology and career. Hundreds of attendees registered to get their curiosities quenched by some of the leading experts in the field including Viral Shah, Mathangi Sri, Abhinanda Sarkar, Sundara Ramalingam N and more.
The other big highlight were the hackathons for machine learning and data science enthusiasts (i.e. MLDS Hackathon and Codeblitz) conducted by Analytics India Magazine’s MachineHack team. The problem statement of MLDS Hackathon was Drones: StatLog Prediction to classify the status of Drone, based on the multiple such instances of Drone Flights. Various attributes of Drone in flight were recorded and the features were anonymised for information copyright concerns. Winners from the two hackathons took away bounties worth more than ₹50,000 in total.
Adding to so many exciting things at MLDS 2020, Analytics India Magazine presented 40Under40 awards, recognising India’s young Data Scientists and leaders who are supporting the growth of analytics in their organisation, have deep industry and analytics expertise and are addressing the insights gap for clients by addressing their most complex challenges.
Stay tuned for the highlights from MLDS 2020 Round 2 which will take place in Hyderabad on 30 and 31 January.