Finding Your Audience In The Internet of Things
summary
Automatic Labs collects OBD data and compresses trips into velocity-acceleration frequency histograms, denoises them with a stacked denoising autoencoder, then clusters using non-negative matrix factorization (NMF) with Kullback–Leibler divergence. Results produce interpretable latent factors — e.g. “serene highway” vs. aggressive city/taxi profiles — and power a public web demo and a product that tags users for targeted fuel-saving audio prompts. The same pipeline applied to velocity vs. RPM reveals gear/torque patterns and flags hybrids/CVTs.
AI-written summary · about
chapters (4)
related talks
01h 09mIoT Realized - The Connected Car
21mIoT, Wavelets and Machine Learning in the Smart Home
01h 15mIoT Realized - The Connected Car v2
22mBuilding a movement data and analytics platform to de-stress customers & save lives
06mWhen Sharing Economy Meets IoT: Towards Fine-grained Urban Air quality Monitoring through Mobile ...
19mTyler Bell | Google Next 2017
20mAWS On Air ft. AWS IoT FleetWise
37mFrom Sensors to Insights at the Edge: Applying AI to IoT Data