In the this second post, we conclude our exercise of builiing a neural net from scratch. We implement backpropagation, make predictions, test the accuracy of the model using various performance metrics, and compare our neural net with a logistic regression model.
In this Two-part series, we will build a shallow neural net from scratch and see how it compares with a logistic regression model. In this first part, we present the dataset we are going to use, the pre-processing involved, the train-test split, and describe in detail the architecture of the model.
This post examines design considerations in integrating standard and portable C++ code in an R package, using Rcpp in the interface level alone.
This is the first in a series of post looking at the best practice of writing Rcpp applications that keep both reusable and standard C++ code separate from the R interface.
In earlier posts, we shared how we wrote our book Data Science in Education Using R as an open book. In this post, we describe what we consider to be the open-source authorship.
One hundred eighty-four new packages stuck to CRAN in May. This post lists my Top 40 picks in eleven categories.
This post describes how to make R notify you via email, text, Slack, or Teams, when a long-running task completes or a ‘fire alarm’ is pulled.
People learning R for work have to imagine how they might apply what they are learning to the kinds of problems the see on their jobs. This post explains the concept and shows how to take the first steps.
This post highlights three outstanding COVID-19 resources: an illuminating dashboard, the COVID-19 Data Hub, and the four talks given at the recent COVID-19 Data Forum webinar.
One hundred forty-eight new packages made it to CRAN in April. Here are my Top 40 picks in nine categories: Computational Methods, Data, Machine Learning, Medicine, Science, Statistics, Time Series, Utilities, and Visualization.