32 research idea notes tagged “methods”. Read them as Markdown, JSON or an Atom feed.
Digital leadership appears on 17 recent papers and leadership keywords on almost 300, and 70 leadership papers carry a digital or AI keyword. New leadership constructs keep appearing, each with its own survey scale. Is digital leadership something new, or transformational leadership measured with new words?
Papers in the last 30 days — leadership: 4
The synthetic control method appears on 29 recent papers, mostly in economics, and its use is falling (4 in the last quarter against about 8 expected), while difference-in-differences is the most-used method on this site. Do published synthetic control studies report what readers need to trust them: pre-period fit, donor pools and placebo tests?
Papers in the last 30 days — synthetic control method: 1
Qualitative methods appear on 46 recent papers, mostly in management, and only 7 of them carry an AI or language-model keyword. Researchers are starting to let language models code interviews and open-ended answers. Do the codes agree with trained human coders, and what does the method section need to report?
Papers in the last 30 days — fuzzy-set qualitative comparative analysis: 2
Inference latency and resource-constrained deployment are among the fastest-growing keywords in the AI journals, yet only 15 of 226 such papers carry an energy, carbon or green-AI keyword. How much computing do the methods in AI journals cost to train and run, and would reporting it change which methods win?
Papers in the last 30 days — inference latency: 4 · resource-constrained deployment: 2
Tabular data appears on 53 recent papers and tabular classification on 13, almost all in AI journals, and 16 tabular papers involve pretrained or language models. Most business data is tables: loans, transactions, patients, sales. Do pretrained tabular models beat gradient boosting on real business tables, and when?
Papers in the last 30 days — tabular data: 4 · tabular data classification: 1
Diffusion models are among the largest topics in the AI journals, and synthetic data generation is growing with them. Yet only 6 of 119 recent synthetic-data papers carry a finance keyword. Could synthetic transaction and loan data let banks share data for research without exposing customers, and how close is it to the real thing?
Papers in the last 30 days — synthetic data generation: 3 · diffusion models: 18
Knowledge distillation, model compression and small language models appear on more than 700 recent papers in AI journals, and only 12 of them concern finance, banking, legal or business tasks. Banks cannot send customer data to the largest models, and regulators want models they can inspect. How small can a model be and still do financial work, and what is lost in the shrinking?
Papers in the last 30 days — knowledge distillation: 23 · small language models: 1
Research integrity, plagiarism, research misconduct and retractions appear on 28 recent papers across the AI, finance and economics journals. Retractions have reached record numbers, and paper mills sell authorship and citations, with language models making fabrication cheaper. How much of the recent literature in these fields is affected, and can it be detected?
Papers in the last 30 days — research integrity: 2 · research misconduct: 1
Quantum computing, quantum machine learning and quantum annealing appear on 115 recent papers, mostly in AI journals, and 33 of them concern optimization. Firms in finance and logistics run quantum pilots. Do the quantum results beat the best classical solvers on the same instances, or only the classical baselines the papers chose?
Papers in the last 30 days — quantum computing: 1 · quantum machine learning: 2
Continuous-time models built on the Hamilton-Jacobi-Bellman equation are rising in the finance journals, while neural operators and physics-informed networks that solve such equations are growing in AI. All five recent papers that pair the two are in AI journals. Can these solvers make richer portfolio and asset-pricing models usable?
Papers in the last 30 days — Hamilton-Jacobi-Bellman equation: 1 · physics-informed neural networks: 8
"Chinese listed companies" is the third most frequent keyword across the tracked journals over the past year, and most of these papers draw on the same few databases. Yet none of the 1,392 recent papers with such a keyword carries a replication, multiple-testing or publication-bias keyword. How many of their findings would survive a correction for everyone testing on the same data?
Papers in the last 30 days — Chinese listed companies: 56 · publication bias: 2
"Moderating effect" and "mediating effect" are among the most frequent keywords in finance and management papers, and mediation models rank high on the equation ranking. How robust are these interaction and mediation results, and what would good practice look like?
Papers in the last 30 days — moderating effect: 20 · mediating effect: 7
Synthetic data generation appears on 77 recent papers and synthetic data on 18, almost all in AI journals, while model collapse, the degradation of models trained on their predecessors' output, appears on 3. The web is filling with generated text and images that will train the next models. How fast does quality decay, and what does that do to the value of human data?
Papers in the last 30 days — synthetic data: 1
Confidence calibration and uncertainty calibration appear on 54 recent papers, mostly in AI journals, and 35 calibration papers involve language or vision-language models, with 94 calibration-related papers in the last quarter against about 57 per quarter before. When a model answers a financial or legal question, its stated confidence rarely matches its accuracy. Can calibration be fixed, and do users read it?
Papers in the last 30 days — uncertainty calibration: 1
Economic, climate, trade and monetary policy uncertainty indices appear on 423 recent papers in the finance and economics journals, almost all built from counts of newspaper words. Only 5 of them use language models, embeddings or text mining beyond word counts. Would an index that reads the articles agree with the word counts, and would it predict better?
Papers in the last 30 days — economic policy uncertainty: 5 · text mining: 1
Evaluation metrics appear on 56 recent papers and LLM-as-a-judge on 16, all in AI journals, with 25 evaluation papers in the last quarter in line with earlier quarters. Models now grade other models' answers, and firms use them to score customer conversations and reports. How reliable and how biased are the model judges, and how far do the biases propagate?
Papers in the last 30 days — evaluation metrics: 4 · LLM evaluation: 2
Benders decomposition, column generation, branch-and-cut and related methods appear on 240 recent papers, almost all in the operations research journals, and the topic is fading, with 42 papers in the last quarter against about 64 per quarter before. Only 16 of them use learning. Can learned cut selection, column pricing and branching make these methods faster on real problems?
Papers in the last 30 days — Benders decomposition: 4 · column generation: 3
Survey experiments are a staple of economics, and some researchers now simulate respondents with language models ("silicon samples"). Only 6 of 179 recent survey papers carry a language-model keyword, 5 of them in AI journals. Where do simulated answers match real people, and where would they mislead economic research?
Papers in the last 30 days — survey experiment: 1 · large language models: 95
Matheuristics and local search are among the fastest-growing keywords in the operations research journals, and recent systems let language models write and evolve heuristic code. Only 2 of the 252 heuristics papers in OR journals carry a language-model keyword. Can generated heuristics beat hand-designed ones on standard routing and scheduling instances?
Papers in the last 30 days — matheuristic: 3 · local search: 3
The technology acceptance model is one of the fastest-growing keywords across the tracked journals, and a quarter of recent TAM and UTAUT papers now study AI. But the model was built for workplace software decades ago. Do perceived usefulness and ease of use predict who actually uses generative AI, and what do they miss?
Papers in the last 30 days — technology acceptance model: 4 · UTAUT: 1
Difference-in-differences is the most-mentioned method on the equation ranking, and treatment effect heterogeneity is one of the fastest-rising keywords in economics. A first check finds that about three in ten recent DiD papers with searchable full text name a heterogeneity-robust estimator. How many staggered-adoption findings would survive one?
Papers in the last 30 days — staggered difference-in-differences: 1
The equity-efficiency trade-off and fair division appear on 49 recent papers across the economics, operations research and AI journals. Routing, scheduling and allocation models increasingly add fairness constraints, for drivers, patients or regions. How much efficiency do the constraints cost in practice, and do the beneficiaries see them as fair?
Papers in the last 30 days — fair division: 2 · equity-efficiency trade-off: 2
In the operations research journals, 1,178 recent papers carry an optimization keyword, and 36 of them carry an explanation or interpretability keyword, while interpretability appears on 379 papers in the archive, almost all about machine learning. Planners reject optimized schedules they cannot understand. What counts as an explanation of a solution, and do explanations raise adoption?
Papers in the last 30 days — decision support system: 4 · interpretability: 17
Class imbalance is one of the largest topics in the AI journals, and fraud detection is where it matters most for finance. But only 10 of 174 recent fraud papers carry a cost, profit or loss keyword; most report accuracy or F1 scores. Would ranking fraud models by the money they save change which ones win?
Papers in the last 30 days — fraud detection: 3 · class imbalance: 22
Supplier selection is one of the fastest-growing keywords in the operations research journals, and much of it relies on multi-criteria methods such as AHP, TOPSIS and the best-worst method. These methods can rank the same options differently. How often would a paper's conclusion change with the method, and what should a robustness check look like?
Papers in the last 30 days — supplier selection: 6 · TOPSIS: 1
Explainable AI is one of the largest topics in the AI journals and is spreading into management, finance and economics. Explanation fidelity (whether an explanation reflects what the model really does) is a fast-rising but still small keyword. When SHAP, LIME and attention maps disagree about the same prediction, which should a loan officer or a doctor believe?
Papers in the last 30 days — explainable AI: 29 · explanation fidelity: 1
Reinforcement learning appears on more than 700 recent papers, and 41 of them apply it to portfolios, trading or hedging, almost all in AI and operations research journals. The results are rarely reproduced and rarely compared with simple rules after costs. Which of the reported gains survive replication on new data?
Papers in the last 30 days — deep reinforcement learning: 17 · portfolio optimization: 13
Data leakage appears on 16 recent papers and memorization on 3, almost all in AI journals, while thousands of papers report results on public benchmarks that language models may have seen in training. How much of reported progress is contamination, and can finance and economics benchmarks be protected?
Papers in the last 30 days — data leakage: 2
Distribution shift, concept drift and non-stationary time series appear on 371 recent papers, almost all in AI journals, and only 27 of them concern credit, fraud, banking or markets. Credit and fraud models decay as borrowers, fraudsters and the economy change. How fast, how should it be detected, and when should a model be retrained rather than recalibrated?
Papers in the last 30 days — concept drift: 6 · distribution shift: 8
Double machine learning, causal forests, conditional average treatment effects and policy learning appear on 50 recent papers, split between the AI and economics journals, with 14 in the last quarter against about 11 per quarter before. The methods promise to find who benefits most from a programme. Do the targeting rules they produce hold up in new samples, and are they fair?
Papers in the last 30 days — double machine learning: 2
Distributionally robust optimization is one of the fastest-growing methods in operations research, and uncertainty estimation is rising in the AI journals, yet only two papers in the archive touch both. Can calibrated uncertainty from machine learning decide how much robustness an optimization model buys?
Papers in the last 30 days — distributionally robust optimization: 9 · conformal prediction: 2
The dataset ranking shows CIFAR and MNIST still among the most-mentioned image benchmarks in AI journal papers, while cross-dataset generalization and reproducibility are rising keywords. Do gains on these old benchmarks carry over to harder, shifted data?
Papers in the last 30 days — benchmark dataset: 6 · cross-dataset generalization: 2