📢📡 AI for All, Zoho Analytics & Tokyo Olympics 🏁✌

The Belamy | Weekly stories from the world of AI

We have lined up exciting online events on data science & AI for the next 2 months. Go register.

Masterclass | Do You Think You Can Analytics? | 10th Aug 2021


Hands-on Workshop | Accelerate PyTorch Applications | 18th Aug 2021


Virtual Conference | Rakuten Applied AI Conference 2021 | 19-20th Aug 2021


Hands-On Workshop | Mastering Exploratory Data Analysis | 28th Aug 2021


Virtual Conference | Deep Learning DevCon 2021 | 23-24th Sep 2021

Early bird passes expiring on Friday



PM Modi Launches AI Initiatives

Indian Prime Minister Narendra Modi has launched key initiatives marking the first anniversary of National Education Policy (NEP 2020). 

The major announcements include the AI for All initiative by the central board of secondary education(CBSE) in collaboration with Intel, SAFAL assessment framework etc. AI For All is a four-hour, self-paced learning program that demystifies AI in an inclusive manner.


What Makes Zoho Analytics Stand Apart?

Zoho Analytics is a business intelligence (BI) and analytics software, first rolled out in 2009. It offers cutting edge analytics solutions to transform raw data into actionable insights.

The platform lets the user fetch data from any data source and analyse it visually to make informed and data-driven decisions. It also allows the user to easily share insights and collaborate.


How Tokyo Olympics Is Using Tech

The Tokyo Olympics 2020 is making extensive use of advanced technology for effective management–from robot assistance to immersive live viewing. In addition, athletes are employing analytics and AI to train and track their performance.

Toyota Motor Corp is playing A crucial role in bringing robots to the Tokyo Olympics in a bid to reduce the number of human volunteers in the light of COVID-19 pandemic. The AI-powered self-driving Field Support Robots (FSR) equipped with cameras and sensors are used to retrieve items like javelins.


Data Science Hiring Process At Rapido

Rapido has nearly 35 people in its data science team in various roles, including data scientists, analysts, data engineers and product managers with experience in data-led and experiment-driven product building.

Rapido is now looking to create a second generation of data products, betting on five critical growth areas with high optimisation and algorithmic potential. The company is also expanding its data science team in Bengaluru.


Featured Video | Heike E. Riel, IBM Fellow, Head Science & Technology, Lead IBM Research Quantum Europe

Heike E. Riel is responsible for leading the research agenda of the Science & Technology department aiming to create scientific and technological breakthroughs.

In an interview with Analytics India Magazine, Heike shared her journey on how she gets into the field of quantum computing, her research work, challenges, and applications of quantum computing. She also has a piece of advice for quantum researchers in India.


Hands-on Guides for ML Developers

Microsoft FLAML VS Traditional ML Algorithms: A Practical Comparison

VIT-AugReg: A PyTorch Image Model Descriptive Predictions

Hands-On Tutorial on Visualizing Spectrograms in Python

Hands-On Guide To Librosa For Handling Audio Files

Complete Guide To SARIMAX in Python for Time Series Modeling



SentinelOne’s Denise Schlesinger On How To Build A Great Data Team

What Separates AI From An Idiot Savant Is Common Sense: Hector Levesque

Tech Behind Hyderabad Based Neobank ZikZuk

Meet The Delhi Based Startup Behind Supreme Court’s AI Portal


Deep Learning On A Data Diet

A common practice is to train small models until they converge and then run a compression technique lightly. Techniques like parameter pruning have already become popular for reducing redundancies without sacrificing accuracy.

Identifying important training data plays a role in online and active learning. But how much of the data is superfluous? Which examples are important for generalisation? And how does one find them?


TensorFlow-Ranking over years

Three years ago, Google introduced TF-Ranking, an open-source TensorFlow-based library for developing scalable neural learning-to-rank (LTR) models. Compared to standard classification models that classify one item at a time, LTR models receive an entire list of items as input and learn an ordering that maximises the utility of the entire list.

In May this year, Google launched the latest version of TF-Ranking that enables full support for natively building LTR models using Keras, a high-level API of TensorFlow 2.



Here's what all happened last week.

NASSCOM Foundation & DXC Technology Launch Employability Skills Program

INSOFE Launches Dual Specialisation Master’s In Computer Science & AI

Conversational AI Firm Uniphore To Acquire Jacada

Tredence To Hire Over 1,000 Analytics & Engineering Professionals By 2022

IIT Jodhpur To Offer B.Tech In AI & Data Science