Notice: This website is continuously updated as new conference proceedings and journal volumes are published. Consequently, the figures presented here may differ from those reported in our publications.
A brief overview of our methodology is provided on the Methods page, along with references to publications documenting our approach and links to the code and data used in each study.
The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates.
Learning Datum-Wise Sampling Frequency for Energy-Efficient Human Activity Recognition
Authors: Weihao Cheng, Sarah Erfani, Rui Zhang, Ramamohanarao Kotagiri
AAAI 2018 | Venue PDF | LLM Run Details | Input Tokens: 21,920 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 4,007 Total number of tokens produced by the LLM (including reasoning/thinking tokens) for this paper's analysis.
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate DWFS with three real-world HAR datasets, and the results show that DWFS statistically outperforms the state-of-the-arts regarding a combined measurement of accuracy and energy efficiency. |
| Researcher Affiliation | Academia | School of Computing and Information Systems, The University of Melbourne {weihaoc@student.,sarah.erfani@,rui.zhang@,kotagiri@}unimelb.edu.au |
| Pseudocode | Yes | Algorithm 1 DWFS Generate H; Algorithm 2 DWFS Learning; Algorithm 3 DWFS Inference |
| Open Source Code | No | The paper does not provide any explicit statements or links indicating that the source code for the described methodology is open-source or publicly available. |
| Open Datasets | Yes | Human Activity Sensing Consortium (HASC) 2011 (Kawaguchi et al. 2011): The data of 6 activities was collected with 100 readings per second by 7 subjects using i Phone/i Pod. Human Activity Recognition on Smartphones Dataset (HARSD) (Anguita et al. 2013): The data of 6 activities was collected with 50 readings per second by 30 subjects using a Samsung Galaxy S II. Daily Sport Activities dataset (DSA) (Barshan and Y uksek 2013): The data of 19 activities was collected with 25 readings per second by 8 subjects using body-worn sensors (we use the data collected by the sensor placed on a subject s torso). |
| Dataset Splits | Yes | We conduct the experiments based on 5-fold cross-validation, where we take 1/5 of D in turn for testing, and the other 4/5 of D for training. |
| Hardware Specification | No | The paper mentions the operating system ('Python 2.7 on 64-bit Ubuntu 14.04 LTS operating system') but does not specify any hardware components like CPU, GPU models, or memory. |
| Software Dependencies | No | The paper mentions 'Python 2.7 on 64-bit Ubuntu 14.04 LTS operating system' but does not list any other software libraries, frameworks, or solvers with specific version numbers. |
| Experiment Setup | Yes | The iteration round L for training is set to 5. We use softmax regression as the classification model of DWFS, that the parameter θ is an m (d +1) matrix including intercepts, and we optimize θ iteratively by Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm. We model the parameter ψ as a K (d + m) matrix, that ψk(s) = ψk s, and we optimize ψ iteratively via BFGS as well. We set the parameter β = 1.2 which is used to impose instance weight. |