Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 31 of the 33 references it cites were found in a registry or an index; 1 of the 12 sources its claims cite were checked for support.

Activities Recognition, Anomaly Detection and Next Activity Prediction Based on Neural Networks in Smart Homes

Open access
Alaghbari, Khaled A., Md. Saad, Mohamad Hanif, Hussain, Aini, Alam, Muhammad Raisul · 2022 · IEEE Access

Abstract

In this paper, we propose a unified deep learning model for monitoring elderly in execution of daily life activities such as eating, sleeping or taking medication. The proposed approach consists of three stages which are activity recognition, anomaly detection and next activity prediction. Such a system can provide useful information for the elderly, caregivers and medical teams to identify activities and generate preventive and corrective measures. In literature, these stages are discussed separately, however, in our approach, we make use of each stage to progress into the next stage. At first, activity recognition based on different extracted features is performed using a deep neural network (DNN), then an overcomplete-deep autoencoder (OCD-AE) is employed to separate the normal from anomalous activities. Finally, a cleaned sequence of consecutive activities is constructed and used by a long short-term memory (LSTM) algorithm to predict the next activity. Since the last two stages depend on the activity recognition stage, we propose to increase its accuracy by exploiting different extracted features. The performance of the proposed unified approach has been evaluated on real smart home datasets to demonstrate its ability to recognize activities, detect anomalies and predict the next activity.

Overview

CoChat Summary

TL;DR

The paper proposes a unified deep learning pipeline for smart homes that performs activity recognition with a DNN, anomaly detection with an overcomplete-deep autoencoder, and next activity prediction with LSTM. It evaluates the approach on the CASAS Aruba and Cairo datasets, reporting 93% and 76% activity-recognition accuracy, respectively, and LSTM next-activity accuracy of 54.6% on Aruba and 45.4% on Cairo.

Key findings

  • Activity recognition: the proposed DNN achieved 93% accuracy on Aruba using 42 inputs (duration + sensor count + activities per day + all sensor states) and 76% accuracy on Cairo using 35 inputs; on Aruba, linear SVM achieved the best F1-score at 90% in the combined-feature setting, with DNN competitive.
  • Anomaly detection: the proposed OCD-AE correctly identified the maximum number of true-positive anomalies in 10 out of 11 Aruba activities with accuracy above 90%; on Cairo, OCD-AE performed best in most activities and was outperformed by SVM only for R2 take medicine.
  • Anomaly ground truth based on boxplot outliers yielded 723 anomalies out of 6,477 Aruba activity instances and 37 anomalies out of 600 Cairo instances.
  • Next activity prediction: with sequence length 4 on Aruba, LZ78 achieved 38%, ALZ 50.3%, and LSTM 54.6%; on Cairo, LSTM achieved the best accuracy at 45.4%.

Methods

  • Design: a three-stage unified model where DNN activity recognition feeds an overcomplete-deep autoencoder for anomaly detection, after which anomalous instances are discarded and the cleaned activity sequence is used by LSTM to predict the next activity.
  • Data: CASAS Aruba (39 sensors, 220 days, 11 activities, 6,477 activity instances, 1,719,533 raw sensor readings) and Cairo (27 motion sensors plus 5 temperature sensors, 57 days, 13 activities, 600 activity instances, 726,534 raw sensor readings); features were duration, sensor count, activities per day, and sensor states, with a 70:30 training-to-test split.
  • Analysis: activity recognition compared DNN against decision tree, random forest, KNN, SVM, and Naive Bayes; anomaly detection compared OCD-AE against SVM, K-means, and an H2O autoencoder; sequence prediction compared LSTM against LZ78 and ALZ using accuracy, precision, recall, F1-score, and MCC.

Limitations

  • Stated: the autoencoder requires manually selecting a threshold value based on reconstruction error to classify anomaly events; sequence prediction accuracy is low due to the limited length of sequence used for training; future work should handle the imbalanced dataset to improve deep learning models.
  • Stated: there is no publicly available dataset on abnormal behaviors for people suffering from dementia, so the paper generates approximate ground truth using boxplot outlier analysis rather than direct observation.
  • Evident from the text: anomaly labels are heuristic (duration and sub-event outliers), class imbalance is present, and Cairo has very few anomaly instances (37 out of 600), which may limit anomaly-detection performance and generalization; Aruba’s Resperate activity has only 6 instances and no recorded anomalies.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1109/access.2022.3157726

Cites

31 of 31 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

1 Valid · 1 Not valid · 9 Not checked · 51 Own results and interpretations

28 more claims

Create a free account to see all 62 claims and to verify the 2 references still open.

See all claims — free account
  • It may state two separate claims as one. 17
  • Probably not a substantial claim. 13
  • The quote only partly states this claim. 2

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

3 of 3

Citations

Cites

Show all 33 references

Research works printed but not linked: 2

  1. Ng, 2011, Sparse autoencoder, CS294A Lect. Notes, 72, 1
  2. Sasaki, 2007, The truth of the F-measure, Teach Tutor mater, 1, 1

Cited by

See all 61

Metadata

Authors
Alaghbari, K. A., Md. Saad, M. H., Hussain, A., & Alam, M. R.
Affiliations
  • National University of Malaysia (MY)
  • Institute of IR4.0, Universiti Kebangsaan Malaysia (UKM), Bangi, Selangor, Malaysia
  • University of Toronto (CA)
Venue
IEEE Access
Published
2022
Volume
10
Pages
28219-28232
ISSN
2169-3536 (ISSN-L)
OpenAlex
W4226247552
Cited by
61
References
33
Citations by year
2026: 12 · 2025: 15 · 2024: 15 · 2023: 14 · 2022: 5
FWCI
4.01
Type
journal article
Language
en
Open access
Locations
  • IEEE Access (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Greater South Information System
Funders
Universiti Kebangsaan Malaysia (DIP-2020-021)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Human Pose and Action Recognition
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › IoT-based Smart Home Systems
Keywords
activity recognition, anomaly detection, next activity prediction, elderly care, home automation, deep autoencoder, long short-term memory, feature extraction, activities of daily living

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Activities Recognition, Anomaly Detection and Next Activity Prediction Based on Neural Networks in Smart Homes. https://app.cochat.ai/papers/10.1109/access.2022.3157726
The paper
Alaghbari, K. A., Md. Saad, M. H., Hussain, A., & Alam, M. R. (2022). Activities Recognition, Anomaly Detection and Next Activity Prediction Based on Neural Networks in Smart Homes. IEEE Access. https://doi.org/10.1109/access.2022.3157726

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.