Back to datasets
Dataset assetClassic DatasetHuman Activity MonitoringActivity Recognition

WISDM Dataset

The WISDM dataset includes six activities: sitting, jogging, walking, descending stairs, ascending stairs, and standing.

Source
github
Created
May 13, 2020
Updated
May 14, 2020
Signals
451 views
Availability
Linked source ready
Overview

Dataset description and usage context

Dataset Overview

Dataset Name

WISDM Dataset (WISDM_ar_v1.1_raw.txt)

Data Source

The dataset is downloaded from http://www.cis.fordham.edu/wisdm/dataset.php

Content

The dataset contains six human activities:

  • Sitting
  • Jogging
  • Walking
  • Descending stairs
  • Ascending stairs
  • Standing

Intended Use

It is used to compare the performance of three models (CNN, LSTM, CNN+LSTM) on human‑activity‑recognition tasks. Performance comparison figures are shown below:

  • Loss comparison chart for the three models: link text
  • Accuracy comparison chart for the three models: link text
Need downstream help?

Pair the dataset with AI analysis and content workflows.

Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.

Explore AI studio