Retail store forecasting with video and audio, ML in high frequency trading, from tech to politics, ML in Web3 - Greg Tanaka, the data scientist show #041

Greg Tanaka is a computer scientist turned CEO of an AI company. He started coding when he was 6, studied computer science at UC Berkeley, and has built many machine learning applications, he is the the founder and CEO of Percolata developing ”Forecast as a Service”. He is also the council member of Palo Alto in California, and just finished his campaign for congress. Today we’ll talk about his career journey, forecasting, machine learning in blockchain and political campaigns. If you like the show subscribe to the channel and give us a 5-star review. Subscribe to Daliana's newsletter on www.dalianaliu.com/ for more on data science. Daliana's LinkedIn: https://www.linkedin.com/in/dalianaliu/ Daliana's Twitter: https://twitter.com/DalianaLiu Greg's Linkedin: https://www.linkedin.com/in/gltanaka/, Twitter: https://twitter.com/GregTanaka Greg's DAO: https://www.gregtanaka.org/dao Highlights: (00:02:10) use computer vision, audio, and Wi-Fi fingerprints to forecast the retail store traffic  (00:21:55) why time series forecast is hard  (00:26:39) how he made the forecasting more stable  (00:28:46) how he troubleshot the spikes and drops in data  (00:36:04) human trading vs algorithmic trading  (00:47:36) his vision of machine learning in blockchain  (00:54:57) why he got into politics  (01:05:57) advises for people who are interested in Web3  (01:11:04) AutoML and the future of machine learning (01:15:36) things he wished he could learn earlier