arXiv subjects

Daniel Campos

Publications and source records attributed to Daniel Campos.

At least 55 records · Page 3Linked to original sources

arXiv2021

Phase transition in non-Markovian animal exploration model with preferential returns

We study a non-Markovian and nonstationary model of animal mobility incorporating both exploration and memory in the form of preferential returns. We derive exact results for the probability of visiting a given number of sites and develop a practical WKB approximation to treat the nonstationary problem. We further show that this model adequately describes empirical movement data of Egyptian fruit bats (Rousettus aegyptiacus) when accounting for inter-individual variation in the population. Finally, we study the probability of visiting any site a given number of times and derive the corresponding mean-field equation. Here, we find a remarkable phase transition occurring at preferential returns which scale linearly with past visits. Following empirical evidence, we suggest that this phase transition reflects a trade-off between extensive and intensive foraging modes.

cond-mat.stat-mech↗

arXiv2021

Informational entropy thresholds as a physical mechanism to explain power-law time distributions in sequential decision-making

While frameworks based on physical grounds (like the Drift-Diffusion Model) have been exhaustively used in psychology and neuroscience to describe perceptual decision-making in humans, analogous approaches for more complex situations like sequential (tree-like) decision making are still absent. For such scenarios, which involve a reflective prospection of future options to reach a decision, we offer a plausible mechanism based on the internal computation of the Shannon's entropy for the different options available to the subjects. When a threshold in the entropy is reached this will trigger the decision, which means that the amount of information that has been gathered through sensory evidence is enough to assess the options accurately. Experimental evidence in favour of this mechanism is provided by exploring human performances during navigation through a maze on the computer screen monitored with the help of eye-trackers. In particular, our analysis allows us to prove that: (i) prospection is effectively being used by humans during such navigation tasks, and a quantification of the level of prospection used is attainable, (ii) the distribution of decision times during the task exhibits power-law tails, a feature that our entropy-based mechanism is able to explain, in contrast to classical decision-making frameworks.

physics.soc-ph↗

arXiv2021

IMG2SMI: Translating Molecular Structure Images to Simplified Molecular-input Line-entry System

Like many scientific fields, new chemistry literature has grown at a staggering pace, with thousands of papers released every month. A large portion of chemistry literature focuses on new molecules and reactions between molecules. Most vital information is conveyed through 2-D images of molecules, representing the underlying molecules or reactions described. In order to ensure reproducible and machine-readable molecule representations, text-based molecule descriptors like SMILES and SELFIES were created. These text-based molecule representations provide molecule generation but are unfortunately rarely present in published literature. In the absence of molecule descriptors, the generation of molecule descriptors from the 2-D images present in the literature is necessary to understand chemistry literature at scale. Successful methods such as Optical Structure Recognition Application (OSRA), and ChemSchematicResolver are able to extract the locations of molecules structures in chemistry papers and infer molecular descriptions and reactions. While effective, existing systems expect chemists to correct outputs, making them unsuitable for unsupervised large-scale data mining. Leveraging the task formulation of image captioning introduced by DECIMER, we introduce IMG2SMI, a model which leverages Deep Residual Networks for image feature extraction and an encoder-decoder Transformer layers for molecule description generation. Unlike previous Neural Network-based systems, IMG2SMI builds around the task of molecule description generation, which enables IMG2SMI to outperform OSRA-based systems by 163% in molecule similarity prediction as measured by the molecular MACCS Fingerprint Tanimoto Similarity. Additionally, to facilitate further research on this task, we release a new molecule prediction dataset. including 81 million molecules for molecule description generation

q-bio.QM↗

arXiv2021

Curriculum learning for language modeling

Language Models like ELMo and BERT have provided robust representations of natural language, which serve as the language understanding component for a diverse range of downstream tasks.Curriculum learning is a method that employs a structured training regime instead, which has been leveraged in computer vision and machine translation to improve model training speed and model performance. While language models have proven transformational for the natural language processing community, these models have proven expensive, energy-intensive, and challenging to train. In this work, we explore the effect of curriculum learning on language model pretraining using various linguistically motivated curricula and evaluate transfer performance on the GLUE Benchmark. Despite a broad variety of training methodologies and experiments we do not find compelling evidence that curriculum learning methods improve language model training.

cs.CL↗

arXiv2021

MS MARCO: Benchmarking Ranking Models in the Large-Data Regime

Evaluation efforts such as TREC, CLEF, NTCIR and FIRE, alongside public leaderboard such as MS MARCO, are intended to encourage research and track our progress, addressing big questions in our field. However, the goal is not simply to identify which run is "best", achieving the top score. The goal is to move the field forward by developing new robust techniques, that work in many different settings, and are adopted in research and practice. This paper uses the MS MARCO and TREC Deep Learning Track as our case study, comparing it to the case of TREC ad hoc ranking in the 1990s. We show how the design of the evaluation effort can encourage or discourage certain outcomes, and raising questions about internal and external validity of results. We provide some analysis of certain pitfalls, and a statement of best practices for avoiding such pitfalls. We summarize the progress of the effort so far, and describe our desired end state of "robust usefulness", along with steps that might be required to get us there.

cs.IR↗

arXiv2021

TREC Deep Learning Track: Reusable Test Collections in the Large Data Regime

The TREC Deep Learning (DL) Track studies ad hoc search in the large data regime, meaning that a large set of human-labeled training data is available. Results so far indicate that the best models with large data may be deep neural networks. This paper supports the reuse of the TREC DL test collections in three ways. First we describe the data sets in detail, documenting clearly and in one place some details that are otherwise scattered in track guidelines, overview papers and in our associated MS MARCO leaderboard pages. We intend this description to make it easy for newcomers to use the TREC DL data. Second, because there is some risk of iteration and selection bias when reusing a data set, we describe the best practices for writing a paper using TREC DL data, without overfitting. We provide some illustrative analysis. Finally we address a number of issues around the TREC DL data, including an analysis of reusability.

cs.IR↗

arXiv2021

Significant Improvements over the State of the Art? A Case Study of the MS MARCO Document Ranking Leaderboard

Leaderboards are a ubiquitous part of modern research in applied machine learning. By design, they sort entries into some linear order, where the top-scoring entry is recognized as the "state of the art" (SOTA). Due to the rapid progress being made in information retrieval today, particularly with neural models, the top entry in a leaderboard is replaced with some regularity. These are touted as improvements in the state of the art. Such pronouncements, however, are almost never qualified with significance testing. In the context of the MS MARCO document ranking leaderboard, we pose a specific question: How do we know if a run is significantly better than the current SOTA? We ask this question against the backdrop of recent IR debates on scale types: in particular, whether commonly used significance tests are even mathematically permissible. Recognizing these potential pitfalls in evaluation methodology, our study proposes an evaluation framework that explicitly treats certain outcomes as distinct and avoids aggregating them into a single-point metric. Empirical analysis of SOTA runs from the MS MARCO document ranking leaderboard reveals insights about how one run can be "significantly better" than another that are obscured by the current official evaluation metric (MRR@100).

cs.IR↗

arXiv2021

Overview of the TREC 2020 deep learning track

This is the second year of the TREC Deep Learning Track, with the goal of studying ad hoc ranking in the large training data regime. We again have a document retrieval task and a passage retrieval task, each with hundreds of thousands of human-labeled training queries. We evaluate using single-shot TREC-style evaluation, to give us a picture of which ranking methods work best when large data is available, with much more comprehensive relevance labeling on the small number of test queries. This year we have further evidence that rankers with BERT-style pretraining outperform other rankers in the large data regime.

cs.IR↗

arXiv2020

Continuous time random walks under Markovian resetting

We investigate the effects of markovian resseting events on continuous time random walks where the waiting times and the jump lengths are random variables distributed according to power law probability density functions. We prove the existence of a non-equilibrium stationary state and finite mean first arrival time. However, the existence of an optimum reset rate is conditioned to a specific relationship between the exponents of both power law tails. We also investigate the search efficiency by finding the optimal random walk which minimizes the mean first arrival time in terms of the reset rate, the distance of the initial position to the target and the characteristic transport exponents.

cond-mat.stat-mech↗

arXiv2020

ORCAS: 18 Million Clicked Query-Document Pairs for Analyzing Search

Users of Web search engines reveal their information needs through queries and clicks, making click logs a useful asset for information retrieval. However, click logs have not been publicly released for academic use, because they can be too revealing of personally or commercially sensitive information. This paper describes a click data release related to the TREC Deep Learning Track document corpus. After aggregation and filtering, including a k-anonymity requirement, we find 1.4 million of the TREC DL URLs have 18 million connections to 10 million distinct queries. Our dataset of these queries and connections to TREC documents is of similar size to proprietary datasets used in previous papers on query mining and ranking. We perform some preliminary experiments using the click data to augment the TREC DL training data, offering by comparison: 28x more queries, with 49x more connections to 4.4x more URLs in the corpus. We present a description of the dataset's generation process, characteristics, use in ranking and suggest other potential uses.

cs.IR↗

arXiv2020

XGLUE: A New Benchmark Dataset for Cross-lingual Pre-training, Understanding and Generation

In this paper, we introduce XGLUE, a new benchmark dataset that can be used to train large-scale cross-lingual pre-trained models using multilingual and bilingual corpora and evaluate their performance across a diverse set of cross-lingual tasks. Comparing to GLUE(Wang et al., 2019), which is labeled in English for natural language understanding tasks only, XGLUE has two main advantages: (1) it provides 11 diversified tasks that cover both natural language understanding and generation scenarios; (2) for each task, it provides labeled data in multiple languages. We extend a recent cross-lingual pre-trained model Unicoder(Huang et al., 2019) to cover both understanding and generation tasks, which is evaluated on XGLUE as a strong baseline. We also evaluate the base versions (12-layer) of Multilingual BERT, XLM and XLM-R for comparison.

cs.CL↗

arXiv2020

On the Reliability of Test Collections for Evaluating Systems of Different Types

As deep learning based models are increasingly being used for information retrieval (IR), a major challenge is to ensure the availability of test collections for measuring their quality. Test collections are generated based on pooling results of various retrieval systems, but until recently this did not include deep learning systems. This raises a major challenge for reusable evaluation: Since deep learning based models use external resources (e.g. word embeddings) and advanced representations as opposed to traditional methods that are mainly based on lexical similarity, they may return different types of relevant document that were not identified in the original pooling. If so, test collections constructed using traditional methods are likely to lead to biased and unfair evaluation results for deep learning (neural) systems. This paper uses simulated pooling to test the fairness and reusability of test collections, showing that pooling based on traditional systems only can lead to biased evaluation of deep learning systems.

cs.IR↗

arXiv2020

Overview of the TREC 2019 deep learning track

The Deep Learning Track is a new track for TREC 2019, with the goal of studying ad hoc ranking in a large data regime. It is the first track with large human-labeled training sets, introducing two sets corresponding to two tasks, each with rigorous TREC-style blind evaluation and reusable test sets. The document retrieval task has a corpus of 3.2 million documents with 367 thousand training queries, for which we generate a reusable test set of 43 queries. The passage retrieval task has a corpus of 8.8 million passages with 503 thousand training queries, for which we generate a reusable test set of 43 queries. This year 15 groups submitted a total of 75 runs, using various combinations of deep learning, transfer learning and traditional IR ranking methods. Deep learning runs significantly outperformed traditional IR runs. Possible explanations for this result are that we introduced large training data and we included deep models trained on such data in our judging pools, whereas some past studies did not have such training data or pooling.

cs.IR↗

arXiv2019

Optimal management of impaired self-avoiding random walks for minimizing spatial coverage

Self-avoidance is a common mechanism to improve the efficiency of a random walker for covering a spatial domain. However, how this efficiency decreases when self-avoidance is impaired or limited by other processes has remained largely unexplored. Here we use simulations to study the case when the self-avoiding signal left by a walker both (i) saturates after successive revisits to a site, and (ii) evaporates, or dissappears, after some characteristic time. We surprisingly reveal that the mean cover time becomes minimum for intermediate values of the evaporation time, leading to the existence of a nontrivial optimum management of the self-avoiding signal. We argue that this is a consequence of complex blocking effects caused by the interplay with the signal saturation and, remarkably, we show that the optimum becomes more and more significant as the domain size increases.

cond-mat.stat-mech↗

arXiv2019

Universal scaling in bidirectional flows of self-avoiding agents

The analysis of the radial distribution function of a system provides a possible procedure for uncovering interaction rules between individuals out of collective movement patterns. This approach from classical statistical mechanics has revealed recently the existence of a universal scaling in systems of pedestrians, provided the potential of interaction $V(τ)$ is conveniently defined in the space of the times-to-collision $τ$ [Phys. Rev. Lett. \textbf{113}, 238701 (2014)]. Here we significantly extend this result by comparing numerically the performance of completely different rules of self-avoidance in bidirectional systems and proving that all of them collapse to a common scaling both in the disordered phase ($V(τ) \sim τ^{-2}$) and in the lane-formation regime ($V(τ) \sim τ^{-1}$), so suggesting that these scalings represent actually a universal feature of any self-avoiding bidirectional flow.

cond-mat.soft↗

arXiv2019

Open Domain Web Keyphrase Extraction Beyond Language Modeling

This paper studies keyphrase extraction in real-world scenarios where documents are from diverse domains and have variant content quality. We curate and release OpenKP, a large scale open domain keyphrase extraction dataset with near one hundred thousand web documents and expert keyphrase annotations. To handle the variations of domain and content quality, we develop BLING-KPE, a neural keyphrase extraction model that goes beyond language understanding using visual presentations of documents and weak supervision from search queries. Experimental results on OpenKP confirm the effectiveness of BLING-KPE and the contributions of its neural architecture, visual features, and search log weak supervision. Zero-shot evaluations on DUC-2001 demonstrate the improved generalization ability of learning from the open domain data compared to a specific domain.

cs.CL↗

arXiv2019

Transport properties of random walks under stochastic non-instantaneous resetting

Random walks with stochastic resetting provides a treatable framework to study interesting features about central-place motion. In this work, we introduce non-instantaneous resetting as a two-state model being a combination of an exploring state where the walker moves randomly according to a propagator and a returning state where the walker performs a ballistic motion with constant velocity towards the origin. We study the emerging transport properties for two types of reset time probability density functions (PDFs): exponential and Pareto. In the first case, we find the stationary distribution and a general expression for the stationary mean square displacement (MSD) in terms of the propagator. We find that the stationary MSD may increase, decrease or remain constant with the returning velocity. This depends on the moments of the propagator. Regarding the Pareto resetting PDF we also study the stationary distribution and the asymptotic scaling of the MSD for diffusive motion. In this case, we see that the resetting modifies the transport regime, making the overall transport sub-diffusive and even reaching a stationary MSD., i.e., a stochastic localization. This phenomena is also observed in diffusion under instantaneous Pareto resetting. We check the main results with stochastic simulations of the process.

cond-mat.stat-mech↗

arXiv2019

Anomalous diffusion in random-walks with memory-induced relocations

In this minireview we present the main results regarding the transport properties of stochastic movement with relocations to known positions. To do so, we formulate the problem in a general manner to see several cases extensively studied during the last years as particular situations within a framework of random walks with memory. We focus on (i) stochastic motion with resets to its initial position followed by a waiting period, and (ii) diffusive motion with memory-driven relocations to previously visited positions. For both of them we show how the overall transport regime may be actively modified by the details of the relocation mechanism.

cond-mat.stat-mech↗