hp2500 Goto Github PK

adv_python_quiz2

aimodelshare

automlbenchmark

OpenML AutoML Benchmarking Framework

dabl

Data Analysis Baseline Library

dabl_portfolio_classifier

Running machine learning experiments to create a portfolio classifier for dabl

linkedin

mars

Demos of a game-based intelligence assessment implemented in Minecraft / Project Malmo

mars_minecraft

Video games are a promising tool for the psychometric assessment of cognitive abilities. They can present novel task types and answer formats, they can record process data, and they can be highly motivating for test takers. This paper introduces the first game-based intelligence assessment implemented in Minecraft, an exceptionally popular video game with 176m copies sold. A matrix-based pattern completion task (PC), a mental rotation task (MR) and a spatial construction task (SC) were implemented in the three-dimensional, interactive environment of the game. PC was intended as a measure of inductive reasoning, whereas MR and SC were measures of spatial ability. We tested 129 children aged 10-12 years old on the Minecraft-based tests as well as equivalent pen-and-paper tests. All three scales fit the Rasch model and were moderately reliable. Factorial validity was good with regard to the distinction between PC and SC, but no distinct factor was found for MR. Convergent validity was good as abilities measured with Minecraft and conventional tests were highly correlated at the latent level (r= .72). Subtest-level correlations were in the moderate range. Furthermore we found that behavioral log-data collected from the game environment was highly predictive of performance in the Minecraft test and, to a lesser extent, also predicted scores in conventional tests. The results indicate that Minecraft is a suitable platform for game-based intelligence assessment and encourage future work aiming to explore game-based problem solving tasks that would not be feasible on paperor in conventional computer-based tests.

openml-python

Python module to interface with OpenML

regional-personality-covid-19

Social behaviors play a key role in the spread of COVID-19. Consequently, regional variation in personality traits that capture individual differences in these social behaviors may offer new insight into regional variation in COVID-19 cases. Here we combine self-reported personality data (N ≈ 3.5 million people), official COVID-19 prevalence rates, and behavioral observations (N ≈ 29 million people) to show that regional personality differences in the US and Germany predict the regional onset and growth of COVID-19 cases above and beyond a highly conservative set of socio-demographic, socio-economic, and pandemic-related control variables. Openness and Extraversion were consistently related to earlier COVID-19 onsets and steeper growth rates. The opposite pattern was found for Neuroticism. We also shed light on regional responses to the outbreak, showing that regional personality (i) is associated with objective indicators of regional social distancing, and (ii) predicts self-reported social distancing beyond the effects of individual-level personality.

scipy_proceedings

Tools used to generate the SciPy conference proceedings

sklearn-bot

Random bot running sklearn classifiers on OpenML