arXiv subjects

Giovanni Felici

Publications and source records attributed to Giovanni Felici.

11 recordsLinked to original sources

arXiv2025

A Comprehensive Review of Transformer-based language models for Protein Sequence Analysis and Design

The impact of Transformer-based language models has been unprecedented in Natural Language Processing (NLP). The success of such models has also led to their adoption in other fields including bioinformatics. Taking this into account, this paper discusses recent advances in Transformer-based models for protein sequence analysis and design. In this review, we have discussed and analysed a significant number of works pertaining to such applications. These applications encompass gene ontology, functional and structural protein identification, generation of de novo proteins and binding of proteins. We attempt to shed light on the strength and weaknesses of the discussed works to provide a comprehensive insight to readers. Finally, we highlight shortcomings in existing research and explore potential avenues for future developments. We believe that this review will help researchers working in this field to have an overall idea of the state of the art in this field, and to orient their future studies.

cs.LG↗

arXiv2024

A Review on the Applications of Transformer-based language models for Nucleotide Sequence Analysis

In recent times, Transformer-based language models are making quite an impact in the field of natural language processing. As relevant parallels can be drawn between biological sequences and natural languages, the models used in NLP can be easily extended and adapted for various applications in bioinformatics. In this regard, this paper introduces the major developments of Transformer-based models in the recent past in the context of nucleotide sequences. We have reviewed and analysed a large number of application-based papers on this subject, giving evidence of the main characterizing features and to different approaches that may be adopted to customize such powerful computational machines. We have also provided a structured description of the functioning of Transformers, that may enable even first time users to grab the essence of such complex architectures. We believe this review will help the scientific community in understanding the various applications of Transformer-based language models to nucleotide sequences. This work will motivate the readers to build on these methodologies to tackle also various other problems in the field of bioinformatics.

cs.CL↗

arXiv2023

Predicting Transcription Factor Binding Sites using Transformer based Capsule Network

Prediction of binding sites for transcription factors is important to understand how they regulate gene expression and how this regulation can be modulated for therapeutic purposes. Although in the past few years there are significant works addressing this issue, there is still space for improvement. In this regard, a transformer based capsule network viz. DNABERT-Cap is proposed in this work to predict transcription factor binding sites mining ChIP-seq datasets. DNABERT-Cap is a bidirectional encoder pre-trained with large number of genomic DNA sequences, empowered with a capsule layer responsible for the final prediction. The proposed model builds a predictor for transcription factor binding sites using the joint optimisation of features encompassing both bidirectional encoder and capsule layer, along with convolutional and bidirectional long-short term memory layers. To evaluate the efficiency of the proposed approach, we use a benchmark ChIP-seq datasets of five cell lines viz. A549, GM12878, Hep-G2, H1-hESC and Hela, available in the ENCODE repository. The results show that the average area under the receiver operating characteristic curve score exceeds 0.91 for all such five cell lines. DNABERT-Cap is also compared with existing state-of-the-art deep learning based predictors viz. DeepARC, DeepTF, CNN-Zeng and DeepBind, and is seen to outperform them.

q-bio.GN↗

arXiv2023

Optimizing accuracy and diversity: a multi-task approach to forecast combinations

We present a multi-task optimization approach based on a deep learning architecture for time series forecasting. We leverage large collections of time series to identify the weights of forecasting models that can be combined to produce forecasts for each series. This method jointly addresses two tasks: the selection of different forecasting models, and their effective combination. In doing so, it keeps into account, in an original way, both the accuracy and diversity of the forecasting methods. For a given time series, the model combination module extracts features and uses them to optimize the weights of the forecasting methods. Simultaneously, the model selection module extracts other features to identify the subset of methods to be used for the prediction. This selection process is framed as a classification problem, with the labels representing the set of models to be used for a series. These labels are determined by solving an auxiliary optimization problem that identifies accurate and diverse methods for each time series. The outputs of the two modules are then combined and the entire neural network is jointly trained by minimizing a custom loss function via gradient descent optimization. Experimental results on a large set of series from the M4 competition dataset and from real road traffic data show that our proposal enhances point forecast accuracy compared to state-of-the-art methods.

cs.LG↗

arXiv2021

Informed peer review for publication assessments: Are improved impact measures worth the hassle?

In this work we ask whether and to what extent applying a predictor of publications' impact better than early citations, has an effect on the assessment of research performance of individual scientists. Specifically, we measure the total impact of Italian professors in the sciences and economics in a period of time, valuing their publications first by early citations and then by a weighted combination of early citations and impact factor of the hosting journal. As expected, scores and ranks by the two indicators show a very strong correlation, but there occur also significant shifts in many fields, mainly in Economics and statistics, and Mathematics and computer science. The higher the share of uncited professors in a field and the shorter the citation time window, the more recommendable the recourse to the above combination.

cs.DL↗

arXiv2020

Simultaneous Feature Selection and Outlier Detection with Optimality Guarantees

Sparse estimation methods capable of tolerating outliers have been broadly investigated in the last decade. We contribute to this research considering high-dimensional regression problems contaminated by multiple mean-shift outliers which affect both the response and the design matrix. We develop a general framework for this class of problems and propose the use of mixed-integer programming to simultaneously perform feature selection and outlier detection with provably optimal guarantees. We characterize the theoretical properties of our approach, i.e. a necessary and sufficient condition for the robustly strong oracle property, which allows the number of features to exponentially increase with the sample size; the optimal estimation of the parameters; and the breakdown point of the resulting estimates. Moreover, we provide computationally efficient procedures to tune integer constraints and to warm-start the algorithm. We show the superior performance of our proposal compared to existing heuristic methods through numerical simulations and an application investigating the relationships between the human microbiome and childhood obesity.

stat.ME↗

arXiv2019

MIP-BOOST: Efficient and Effective $L_0$ Feature Selection for Linear Regression

Recent advances in mathematical programming have made Mixed Integer Optimization a competitive alternative to popular regularization methods for selecting features in regression problems. The approach exhibits unquestionable foundational appeal and versatility, but also poses important challenges. Here we propose MIP-BOOST, a revision of standard Mixed Integer Programming feature selection that reduces the computational burden of tuning the critical sparsity bound parameter and improves performance in the presence of feature collinearity and of signals that vary in nature and strength. The final outcome is a more efficient and effective $L_0$ Feature Selection method for applications of realistic size and complexity, grounded on rigorous cross-validation tuning and exact optimization of the associated Mixed Integer Program. Computational viability and improved performance in realistic scenarios is achieved through three independent but synergistic proposals.

stat.ME↗

arXiv2019

Predicting long-term publication impact through a combination of early citations and journal impact factor

The ability to predict the long-term impact of a scientific article soon after its publication is of great value towards accurate assessment of research performance. In this work we test the hypothesis that good predictions of long-term citation counts can be obtained through a combination of a publication's early citations and the impact factor of the hosting journal. The test is performed on a corpus of 123,128 WoS publications authored by Italian scientists, using linear regression models. The average accuracy of the prediction is good for citation time windows above two years, decreases for lowly-cited publications, and varies across disciplines. As expected, the role of the impact factor in the combination becomes negligible after only two years from publication.

cs.DL↗

arXiv2016

Revisiting the use of Robust Optimization for optimal energy offering under price uncertainty

We propose a new Robust Optimization method for the energy offering problem of a price-taker generating company that wants to build offering curves for its generation units, in order to maximize its profit while taking into account the uncertainty of market price. Our investigations have been motivated by a critique to another Robust Optimization method proposed in [Baringo and Conejo, 2011], which entails the solution of a sequence of robust optimization problems imposing full protection and defined over a sequence of nested subintervals of market prices: this method presents a number of issues that may severely limit its application and computational efficiency in practice and that may expose a company to the risk of presenting offering curves resulting into suboptimal or even infeasible accepted offers. To tackle all such issues, our method provides for solving one single robust counterpart, considering an intermediate level of protection between null and full protection, and to make energy offers at zero price, practically eliminating the risk of non-acceptance. Computational results on instances provided by our industrial partners show that our new method is able to grant a great improvement in profit.

math.OC↗

arXiv2014

A-priori Upper Bounds for the Set Covering Problem

In this paper we present a new bound obtained with the probabilistic method for the solution of the Set Covering problem with unit costs. The bound is valid for problems of fixed dimension, thus extending previous similar asymptotic results, and it depends only on the number of rows of the coefficient matrix and the row densities. We also consider the particular case of matrices that are \textit{almost} block decomposable, and show how the bound may improve according to the particular decomposition adopted. Such final result may provide interesting indications for comparing different matrix decomposition strategies.

math.CO↗

arXiv2013

Hacking Smart Machines with Smarter Ones: How to Extract Meaningful Data from Machine Learning Classifiers

Machine Learning (ML) algorithms are used to train computers to perform a variety of complex tasks and improve with experience. Computers learn how to recognize patterns, make unintended decisions, or react to a dynamic environment. Certain trained machines may be more effective than others because they are based on more suitable ML algorithms or because they were trained through superior training sets. Although ML algorithms are known and publicly released, training sets may not be reasonably ascertainable and, indeed, may be guarded as trade secrets. While much research has been performed about the privacy of the elements of training sets, in this paper we focus our attention on ML classifiers and on the statistical information that can be unconsciously or maliciously revealed from them. We show that it is possible to infer unexpected but useful information from ML classifiers. In particular, we build a novel meta-classifier and train it to hack other classifiers, obtaining meaningful information about their training sets. This kind of information leakage can be exploited, for example, by a vendor to build more effective classifiers or to simply acquire trade secrets from a competitor's apparatus, potentially violating its intellectual property rights.

cs.CR↗