Great idea to use #splitlearning & #federatedlearning to enable #ML at a large scale while protecting individual data privacy. Huge implications for #healthcare #imaginationinaction @raskarmit

Split learning is a new technique to approach machine learning from sensitive data (that can not be consolidated). Split learning leaves the data in place & uses it #AI #Splitlearning

Advancing Edge Deployments: Solutions for Language Models Optimisation. 🌐 #ModelQuantization: Reduces model size and computational needs. 🔧 Parameter-Efficient #FineTuning: Optimizes specific parameters for efficiency. 🔀 #SplitLearning: Divides workloads between devices and

ML models memorize training data, a major #privacy problem for data collaboration, even w/ #federatedlearning. @gg159136 and @vectorinversion gave a fantastic preso at @opendp_org this morning on applying #differentialprivacy to #splitlearning to solve this. More to come!

AI on Siloed Data: Sharing wisdom not raw data. Beyond anonymization to protect data, towards #datautility. Specific reference to applications in legal. #ImaginationInAction #Splitlearning #datacommons @jamesondempsey

#Software vs #AI: #ArtificialIntelligence is wisdom through example. See #federatedlearning and #SplitLearning in action @TEDxBeaconSt #TEDxBeaconSt #TEDxBeaconStreet @MIT_CSAIL @medialab @IrmaRaste @eViRaHealth @ipfconline1

分散型機械学習のための分割学習に関するワークショップ（SLDML'21）.Workshop on Split Learning for Distributed Machine Learning (SLDML’21).

@sarahookr @sarahookr Few of the Decentralized Compute and AI ecosystems that have code base : . 1. Prime Intellect Diloco - . 2. Exolabs Exo - . 3. MIT Split Learning - . 4. Petals - . 5.

How reliably does #splitlearning safeguard private information?. @gg159136 and @vectorinversion in our @LiveRamp privacy tech group published their paper on #splitnn attacks and thwarting them with #differentialprivacy in @RealAAAI. Check it out:

In his talk at #BAS22, @michikuehne, Founder and CEO of @Katulu_io, outlines the concepts behind #SplitLearning and how it can be transferred to industrial processes across multiple companies: "With federated learning you can solve key problems - without sharing data."

Our latest blogpost features some new perspectives for privacy-preserving #DeepLearning in #healthcare using #encryption .#homomorphicencryption #splitlearning

Safe Paths: A privacy-first approach to contact tracing @medialab @MIT @MILAMontreal #Splitlearning #NeuralNetworks

.@proneat from @MIT talked about #SplitLearning, a recently developed, highly resource efficient method for collaboration in the health sector, allowing to perform distributed deep learning under data sharing constraints. Download slides #datacouncil

Data Council SF '19 Speaker Announcement: @proneat from @MIT will talk about "#SplitLearning: A Resource Efficient Distributed #DeepLearning Method without Sensitive Data Sharing". Book your tickets now #datacouncil

📢Welcome to read the paper by Praveen Joshi et al. | Performance and Information Leakage in Splitfed Learning and Multi-Head #SplitLearning in #HealthcareData and Beyond.👉 #article #machinelearning #datacollection #dataset

How to obtain sufficient high-quality training data for ML models? Collaborative Distributed Machine Learning systems (CDML) could be the ideal solution 👉 #DataScarcity #MachineLearning #FederatedLearning #SwarmLearning #SplitLearning #AssistedLearning

and some of the ongoing MIT projects was really informative and enlightening. Hearing him talk about split learning and federated learning really did open our thinking horizons! (2) .#REDX #splitlearning #federatedlearning #innovation #machinelearning #deeplearning #makermela.

@nesaorg 4/4 Powering the Future With Nesa’s hybrid sharding and privacy co-optimization protocol, we deliver auditable, scalable AI inference. Trusted AI is now on-chain, where privacy and security meet decentralization. #Web3 #AI #DecentralizedAI #ZKML #SplitLearning @nesaorg.

📢#MDPIfutureinternet | Highly Cited Articles in #ArtificialIntelligence. 📝 Multi-Level Split #FederatedLearning for Large-Scale #AIoT System Based on Smart Cities. Citations: 8. #splitlearning #artificialintelligent #edgecomputing. @ComSciMath_Mdpi

#SplitLearning: Distributed #DeepLearning method without sensitive data sharing. SplitNN is a distributed and private deep learning technique to train deep #neuralnetworks over multiple data sources without the need to share raw labelled data directly.