slide deck, test only

Exploratory data analysis of a modest dataset combining tweets, blog posts and news articles suggests how to build a machine learning model to predict the next word typed,a small-screen device like on a smart phone. Since these texts are not a scrambled “bag of words”, the model built must be capable of predicting sequences of words. To help keep the model both accurate and frugal with limited computational resources, the data should be whittled down, transformed into short strings of words, re-coded for punctuation, and filled out with some missing words.

How to get started with the DC Car Crashes map app available at https://drfreedy.shinyapps.io/dc_car_crashes_map_app/

Visiting Washington, DC soon? Despite all the car accidents, stay safe. Don't take Google, Waze or Uber routes without checking the accident-prone spots along the way. Try the DC Car Crashes on your phone, tablet or laptop now.

Visiting Washington, DC soon? Despite all the car accidents, stay safe. Don't take Google, Waze or Uber routes without checking the accident-prone spots along the way. Try the DC Car Crashes on your phone, tablet or laptop now.

An interactive map built on July 10, 2017 by Daniel Friedheim with Leaflet in R. Instructions: Click on a cluster to zoom in, then an icon for details about one of the 22,927 crashes over the 14-month period from 4/1/15 to 5/27/16, including 35 causing fatalities and 1,180 causing major injuries. Source: “DC Car Crashes,” OpenDataDC: http://opendata.dc.gov/datasets/95254fae17bc4792bd47b53f71c2e503_19 .

We built a classification model trained on data from accelerometers on a weight and the belt, arm and forearm of six human subjects to predict whether they executed 10 repetitions of a simple Unilateral Dumbbell Biceps Curl exercise correctly, or not...Only the Linear Discriminant Analysis model appeared to have failed to predict all 20 testing values. Of the remaining three models, the Support Vector Machines model had the fastest runtime, but the Random Forest model was most accurate by a very slim margin with the Generalized Boosted-regression Model the also-ran on both dimensions... [cross validation, oob error, parameter tuning, runtime, machine learning, supervised, caret, randomForest, svmRadial, gbm, lda, algorithm, IoT]

An exploratory data analysis report on National Weather Service Storm Data from 123 Forecast Offices over 19 years.