Recurrence in biological and artificial neural networks

Recurrence is an overloaded term in the context of neural networks, with disparate colloquial meanings in the machine learning and the neuroscience communities. The difference is narrowing, however, as the artificial neural networks (ANNs) used for practical applications are increasingly sophisticated and more like biological neural networks (BNNs) in some ways (yet still vastly different on the whole). In this post we’ll highlight the historic differences in the use of term recurrence within these two communities, highlight some fairly recent deep learning ANN models that creep towards the neuroscience, point to some neuroscience studies that shine light on the function of recurrence, and speculate on future advancements.

Free Book: Classification and Regression In a Weekend

This tutorial began as a series of weekend workshops created by Ajit Jaokar and Dan Howarth. The idea was to work with a specific (longish) program such that we explore as much of it as possible in one weekend. This book is an attempt to take this idea online. The best way to use this book is to work with the Python code as much as you can. The code has comments. But you can extend the comments by the concepts explained here.

Master Your Hypothesis Test

A tutorial on Power, Bootstrapping, Sample Selection, and Outcome Analysis.

Time Series Analysis, Visualization & Forecasting with LSTM

Statistics normality test, Dickey-Fuller test for stationarity, Long short-term memory.

Using Data Cubes with R

Data cubes are a popular way to display multidimensional data. This makes the method suitable for big data. Giving the incredible growth of data it is natural that the method have become increasingly popular. In this article you learn to use R for data cubes.

Understanding Objective Functions in Deep Learning

Data has consumed our day to day lives. The amount of data that’s is available in the web or from other variety of sources is more than enough to get an idea about any entity. The past few years has seen exponential rise in the volume which has resulted into the adaptation of the term Big Data. Most of these generated data are unstructured and could up in any format. Previously computers were not equipped to understand such unstructured data but modern computers coupled with some programs are able to mind such data and extract relevant information from it which has certainly helped many business. Machine Learning is the study of predictive analytics where the structured or unstructured data are analysed and new results are predicted after the model is trained to learn the patterns from historical data. There are several pre-programmed Machine Learning algorithms which helps in building the model and the choice of the algorithm to be used completely depends on the problem statement, the architecture and the relationship among the variables. However, the traditional state-of-the-art Machine Learning algorithms like Support Vector Machines, Logistic Regression, Random Forest, etc., often lacks efficiency when the size of the data increases. This problem is resolved by the advent of Deep Learning which is a sub-field of Machine Learning. The idea behind Deep Learning is more or less akin to our brain. The neural networks in Deep Learning works almost similarly to the neurons in the human brain.

Time Series Forecasting with TensorFlow.js

Pull stock prices from online API and perform predictions using Recurrent Neural Network & Long Short Term Memory (LSTM) with TensorFlow.js framework.

7 Steps to Mastering SQL for Data Science – 2019 Edition

Follow these updated 7 steps to go from SQL data science newbie to practitioner in a hurry. We consider only the necessary concepts and skills, and provide quality resources for each.

Automating Trading and Market Making With Artificial Intelligence

The goal is to capture information in a market’s order books and use that information to predict market movement/direction. That prediction can enable repricing of orders and more efficient market making. Such an approach allows the market maker to provide liquidity whilst making profits at the same time. Market makers are essential to modern markets. They provide the markets with necessary liquidity and make sure the bid/ask spread is reasonably narrow to allow efficient purchasing. This can be taken a step further by market makers that do more than simply provide a constant bidding and asking price. Some market makers trade at higher frequencies and constantly take advantage of inefficiencies as well as small swings in asset prices.

One of the most fascinating advancements in the world of machine learning, is the development of abilities to teach a machine how to understand human communication. This very arm of machine learning is called as Natural Language Processing. This post is an attempt at explaining the basics of Natural Language Processing and how a rapid progress has been made in it with the advancements of deep learning and neural networks.

Recommendation Systems in the Real world

An overview of the process of designing and building a recommendation system pipeline.

PyCharm for Data Scientists

I have recently started using PyCharm as an alternative to Spyder, and am loving it. This article talks about some of the features of PyCharm that made me completely transition to PyCharm from Spyder. The below features are in comparison with Spyder, and not general IDEs.

60+ useful graph visualization libraries

We outline 60+ graph visualization libraries that allow users to build applications to display and interact with network representations of data.

Confidence Intervals in One Picture

Confidence intervals (CIs) tell you how much uncertainty a statistic has. The intervals are connected to confidence levels and the two terms are easily confused, especially if you’re new to statistics. Confidence Intervals in One Picture is an intro to CIs, and explains how each part interacts with margins of error and where the different components come from.

Synchronous Kernels-only Competitions: Real ML in Real Time

We are pleased to share that we now support a general synchronous Kernels-only (KO) format: when you submit a Kernel, Kaggle will run the code against both the public test set and a withheld private test set in real time. To kick things off, you’re invited to join Instant Gratification, our first synchronous Kernels-only competition using our new framework.