In a typical Federated learning paradigm, a connected network exists with a central server node. Each of the nodes trains a local model with data generated at that node and only that model is shared with the server, not the data. In this talk, we will discuss about the decentralized learning approach in federated learning, how it solves data privacy and scalability issues, building deep learning models using federated learning, federated learning in practice with practical use cases and open problems.
Key Takeaways
- Power of decentralized learning
- Data localization, privacy and scalability issues
- Federated Learning in practice
- Deep Neural Network with Federated Learning
- Use cases and open Problems
————————————————————————————————————————————————————
Bio
Arindam Banerjee – Data Scientist | Ericsson
Arindam is experienced in innovative technical consulting, research and development, statistical modelling, data product design and development using various technologies; a Machine Learning expert and Data Scientist with a passion for turning raw data into products, actionable insights, and meaningful stories.
For more than 10 years, Arindam has been associated with MNCs like TCS, Alcatel-Lucent, Ericsson and has been instrumental in designing, building and deploying large-scale data products. In his current role as a Data Scientist he has several international patents filed from Ericsson.
He earned his master’s degree (M.Tech) from Vellore Institute of Technology, Vellore with a specialization in Computer Science and Engineering. He completed his graduation (B.E.) in Electronics and Telecommunication Engineering from Nagpur University.
In his free time, Arindam enjoys playing Guitar and painting.
His areas of interest include Machine Learning, Scalable Software Development, IoT, Statistical Modelling, Wireless Communication etc
Arindam Banerjee – Data Scientist | Ericsson