Biologically-Plausible Learning Algorithms Can Scale to Large Datasets by Will Xiao[1], Honglin Chen[2], Qianli Liao[2] and Tomaso Poggio[2]
[1] Department of Molecular and Cellular Biology, Harvard University
[2] Center for Brains, Minds, and Machines, MIT
Background Consider a layer in a feedforward neural network. Let xi denote the input to the i th neuron in the layer and yj the output of the j th neuron. Let W denote the feedforward weight matrix and Wij the connection between input xi and output y.
KotlinSyft makes it easy for you to **train and execute PySyft models on Android devices**. This allows you to utilize training data located directly on the device itself, bypassing the need to send a user's data to a central server.
[The first use case for our library is here](https://blog.openmined.org/apheris-openmined-pytorch-announcement)
This work aims to derive a non-trivial breakdown point for an algorithm for training a single hiddenlayer Neural Network. In pursuit of this goal, we propose two algorithms for training a network with ReLU activations. The first approach utilizes the partitioning property of the ReLU function while the second approach utilizes the convexity of the activation function.
There is a large disparity in the access to proper education at the school level. We seek to build a problem aware online tutoring system for students to help improve the situation. In this project, we have made a complete solution generator which, given a word problem on arithmetic at the levels of classes 6-8, extracts relevant information from the question and solves the problem in a step-by-step manner. Currently, we are handling only basic speed, time and distance problems.