Arman daliri

Islamic Azad University, Karaj · Iran

Unclaimed · AI-structured profileORCID OpenAlex

ML researcher in Iran: imbalanced-data classification, metaheuristics and ML for healthcare (ECG, liver)

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-09-27) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.

Research profile

Output—

Not yet scored: this profile lacks a verified full publication inventory, so total output, venue quality and long-run productivity cannot be computed. This axis would measure cumulative publication count, publication venues and output trends over the researcher's career.

Impact—

Not yet scored: beyond a raw h-index of 7 and i10-index of 5 from OpenAlex, citation totals and per-paper influence data are insufficient to compute this axis. It would measure citation-based influence and scholarly uptake of the publications.

Focus—

Not yet scored: topic counts exist (imbalanced classification 4, metaheuristics 3, multi-objective optimization 3, ECG 3, autism 3), but venue-level and depth data are too thin to score concentration. This axis would measure focus versus breadth across topics over time.

Collaboration14/100

Evolve scores collaboration 14/100, reflecting an unusually narrow network: OpenAlex attributes only 1 distinct co-author in the recent window, with Hasan Samadbin as the recurring partner. Both 2026 catalog entries are small-team arXiv preprints — the ECG paper with Daliri as first and corresponding author, the liver-disease paper with him in the last author position. The near-solo pattern gives full control but limits independent validation, external data access and co-review, which the low reproducibility ratings partly reflect.

Activity75/100

Evolve scores activity 75/100, consistent with a clearly rising output curve: OpenAlex counts 2 works in 2022, 2 in 2023, 5 in 2024 and 5 in 2025, with 2 already indexed in 2026, and 12 works in the last three years. Much of the recent output appears as arXiv preprints, indicating rapid release of results; the caveat is that volume is currently outpacing methodological depth, as Evolve's critical reviews of both cataloged papers indicate.

Reproducibility—

Not yet scored: no profile-level reproducibility assessment exists yet. The only signals are paper-level — both cataloged 2026 preprints are rated low repro-readiness, one lacking a formal test protocol. This axis would measure code, data and protocol availability for independent replication.

Research summary

Arman Daliri is a machine-learning researcher affiliated with Islamic Azad University, Karaj (Iran). The OpenAlex-derived topic profile centers on imbalanced-data classification (4 works), metaheuristic optimization (3), advanced multi-objective optimization (3), ECG monitoring and analysis (3), and autism spectrum disorder research (3). Bibliometric signals describe an active early-career profile: h-index 7, i10-index 5, 12 works in the last three years, and yearly output rising from 2 (2022) and 2 (2023) to 5 in each of 2024 and 2025, with 2 already indexed in 2026. The collaboration footprint is narrow — 1 distinct co-author, with Hasan Samadbin the recurring partner — producing a 14/100 collaboration score against 75/100 for activity. Two 2026 arXiv preprints are fully reviewed in the Evolve catalog. The Entropy Triangle Method paper (first author, corresponding) reports ~85% sinusoidal/non-sinusoidal ECG classification over 10,646 patients; the liver-disease entry claims reinforcement-learning-based classifier selection lifting Primary Biliary Cirrhosis accuracy from 63% to 98%. Evolve's reviews rate both papers low for reproducibility readiness, citing under-justified selection procedures, absence of formal reinforcement learning, missing test protocols, and overstated claims such as arrhythmia 'prevention'.

Collaboration pattern

The recorded collaboration footprint is minimal: OpenAlex lists 1 distinct co-author in the recent window, and Hasan Samadbin is the sole frequent collaborator, associated with the 2026 arXiv work. Authorship roles alternate — first-and-corresponding on the ECG preprint, last author on the classifier-selection preprint — suggesting a stable small-team workflow rather than a lab group. No international or cross-institutional co-authorship is visible in the snapshot; broader collaboration would likely strengthen external validation of the reported results.

Research keywords

Imbalanced data classificationOversamplingECG classificationCardiac arrhythmia detectionMetaheuristic optimizationMulti-objective optimizationReinforcement learningAlgorithm selectionAutoMLShark Smell OptimizationMedical informaticsCalinski-Harabasz index

Highlights

Entropy Triangle Method: oversampler selection for ECG arrhythmia classification at 10,000-patient scale
First-author, corresponding 2026 arXiv preprint proposing a triangle-region oversampler-selection framework (using Shark Smell Optimization and the Calinski-Harabasz index) reported to reach ~85% sinusoidal/non-sinusoidal classification across 10,646 patients. Evolve's review finds the selection rationale under-justified and the arrhythmia 'prevention' framing overstated; repro-readiness is rated low.
Claimed reinforcement-learning classifier selection for liver-disease prediction (63%→98%)
2026 arXiv preprint (last author) claiming a reinforcement-learning-based selection step raises Primary Biliary Cirrhosis classification accuracy from 63% to 98% on the Mayo Clinic PBC dataset, despite an NAFLD-focused title. Evolve notes the paper contains no formal reinforcement learning and no test protocol, deeming the claim overclaimed; repro-readiness is rated low.

Timeline

2022Earliest year in the OpenAlex snapshot: 2 indexed works.
2024Output steps up to 5 indexed works, doubling the 2 works of 2022–2023.
2025Sustains peak output with 5 more indexed works.
2026Two arXiv preprints enter the Evolve catalog: the Entropy Triangle Method ECG paper (first and corresponding author) and the reinforcement-learning classifier-selection liver paper (last author); 2 works indexed so far this year.

Papers in the Evolve catalog

2026-09-24first authorcorresponding
The Entropy Triangle Method (ETM): A novel framework for the prevention of cardiac arrhythmia with a review of more than 10,000 patients
2026-09-24
Right Choice of Classification Algorithms Based on Reinforcement Learning for Prediction of Non-Alcoholic Fatty Liver

Frequent co-authors

Hasan Samadbin×1

About this profile

This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.

Bibliometrics (h-index 7, i10-index 5, yearly counts, co-author count) come from an OpenAlex snapshot; citation coverage may be incomplete.
Only the 2 cataloged papers have full Evolve reviews; the remaining indexed works are unreviewed and not quality-assessed.
Profile is name-resolved from public bibliographic data (OpenAlex, ORCID 0000-0002-0398-3052); no biography, affiliation history or awards were verified.
Both cataloged papers are 2026 arXiv preprints, not peer-reviewed publications; claims are reported as reviewed, not endorsed.