🚀 An AMAZING event!!! Many thanks to Bojan, Chris and Rohan for joining the panel and special thanks to Anthony for taking the time out of his busy schedule to make a surprise appearance. If you missed it live, you can watch the recording here: https://t.co/vNDUR8y7UR
New competition launched at @Kaggle
🏆Title: Optiver Realized Volatility Prediction
🎯Goal: Apply your data science skills to make financial markets better
💰Total Prize: $100,000
💻Type: Code Competition
📈Count Towards Point/Tiers: Yes
https://t.co/ftQTWRE8LZ
New competition launched at @Kaggle
🏆Title: MLB Player Digital Engagement Forecasting
🎯Goal: Predict fan engagement with baseball player digital content
💰Total Prize: $50,000
💻Type: Code Competition
📈Count Towards Point/Tiers: Yes
https://t.co/ZMsvGqhNf6
New competition launched at @Kaggle
🏆Title: SETI Breakthrough Listen - E.T. Signal Search
🎯Goal: Find extraterrestrial signals in data from deep space
💰Total Prize: $15,000
💻Type: Regular (Non-Code) Competition
📈Count Point/Tiers: Yes
https://t.co/WHF7NhwkSB
Kaggle Notebook: TPS4: Supervised Emphasized Denoising AutoEncoder by @jeongyoonlee.
When you have label data (e.g. pseudo label for the test set, or training DAE only with the training set), we can train both DAE and a classifier simultaneously.
https://t.co/vSRQWtPRzj
Glad that I am not missing out on anything interesting when I still stubbornly default to ADAM: "A Large Batch Optimizer Reality Check: Traditional, Generic Optimizers Suffice Across Batch Sizes" (https://t.co/P7DsR48hYZ)
Have you ever wondered what TPUs are all about? Today we welcome @kierisi, a new Community Advocate, who walks us through how TPUs, their systolic array architecture, and bfloat16 number formats accelerate your deep learning! [WATCH] https://t.co/8z9r94zQum
#tpu#learnwithJesse