Quaderni MOX
Pubblicazioni
del Laboratorio di Modellistica e Calcolo Scientifico MOX. I lavori riguardano prevalentemente il campo dell'analisi numerica, della statistica e della modellistica matematica applicata a problemi di interesse ingegneristico. Il sito del Laboratorio MOX è raggiungibile
all'indirizzo mox.polimi.it
Trovati 1349 prodotti
-
66/2020 - 29/10/2020
Didkovsky, O.; Ivanov, V.; Papini, M.; Longoni, L.; Menafoglio, A.
A comparison between machine learning and functional geostatistics approaches for data-driven analyses of solid transport in a pre-Alpine stream | Abstract | | We address the problem of providing data-driven models for sediment transport in a pre-Alpine stream in Italy. We base our study on a large set of measurements collected from real pebbles, traced along the stream through Radio Frequency IDentificator (RFID) tags after precipitation events. We propose and evaluate two classes of data-driven models -- based on machine learning and functional geostatistics approaches respectively -- to predict the probability of movement of single pebbles within the stream. The first class is built upon gradient boosting decision trees, and allows one to estimate the probability of movement of a pebble, based on the pebbles’ geometrical features, river flow rate, locations, and subdomain types. The second class is built upon functional kriging, a recent geostatistical technique which allows one to predict a functional profile --i.e., the movement probability of a pebble, as a function of the pebbles geometrical features or of the stream's flow rate-- at unsampled locations in the study area.
Although grounded on different perspectives, both these models aim to account for two main sources of uncertainty, namely (i) the complexity of river’s morphological structure, and (ii) the highly-nonlinear dependence between probability of movement, pebble’s size and shape, and the stream’s flow rate.
We extensively compare the performances of the two methods in terms of classification accuracy, and show that, although these techniques are grounded on different perspectives, an overall consistency appears between the methods suggesting that both approaches may provide valuable modeling frameworks for the problem at hand. We finally discuss on the use of the developed models in a bottom-up strategy, which starts with the prediction/classification of a single pebble and then integrates the results into a forecast of the grain-size distribution of mobilized sediments. |
-
64/2020 - 16/10/2020
Fiz, F.; Viganò, L.; Gennaro, N.; Costa, G.; La Bella, L.; Boichuk A.; Cavinato, L.; Sollini, M.; Politi, L. S.; Chiti, A.; Torzilli, G.
Radiomics of Liver Metastases: A Systematic Review | Abstract | | Patients with liver metastases can be scheduled for different therapies
(e.g., chemotherapy, surgery, radiotherapy, and ablation). The choice of the most appropriate treatment should rely on adequate understanding of tumor biology and prediction of survival, but reliable biomarkers are lacking. Radiomics is an innovative approach to medical imaging: it identifies invisible-to-the-human-eye radiological patterns that can predict tumor aggressiveness and patients outcome. We reviewed the available literature to elucidate the role of radiomics in patients with liver metastases. Thirty-two papers were analyzed, mostly (56%) concerning metastases from colorectal cancer. Even if available studies are still preliminary, radiomics provided effective prediction of response to chemotherapy and of survival, allowing more accurate and earlier prediction than standard predictors. Entropy and homogeneity were the radiomic features with the strongest clinical impact. In the next few years, radiomics is expected to give a consistent contribution to the precision medicine approach to patients with liver metastases. |
-
63/2020 - 30/09/2020
Tuveri, M.; Milani, E.; Marchegiani, G.; Landoni, L.; Torresani, E.; Capelli, P.; Sperandio, N.; D’Onofrio, M.; Salvia, R.; Vergara, C.; Bassi, C.
HEMODYNAMICS AND REMODELING OF THE PORTAL CONFLUENCE IN PATIENTS WITH CANCER OF THE PANCREATIC HEAD: A PILOT STUDY | Abstract | | We designed a computational study to evaluate the effects of hemodynamics on portal confluence remodeling in cancer of the pancreatic head. The importance of this study is the finding that altered flow conditions due to tumor growth can disrupt the balance between eutrophic remodeling and degradative process of the vein wall, leading to the complete substitution of the three-layered vein wall and the opportunity to perform a total pancreatectomy with en-bloc resection of the portal confluence. |
-
62/2020 - 14/09/2020
Massi, M. C.; Ieva, F.
Representation Learning Methods for EEG Cross-Subject Channel Selection and Trial Classification | Abstract | | EEG is a non-invasive powerful system that finds applications in several domains and research areas. At the moment, most EEG systems require subjects to wear several electrodes on the scalp. However, a large number of channels might include noisy information, redundant signals, induce longer preparation times and increase the computational times of any automated system trying to classify EEG recordings. One way to reduce the signal-to-noise ratio and improve the classification accuracy is to combine channel selection with feature extraction. However, when dealing with EEG channel selection most of the efforts have been focused on identifying the most relevant channels in a subject-dependent fashion. In this paper we introduce a novel algorithm for subject independent channel selection of EEG recordings.
In particular, the algorithm (i) exploits channel-specific Representation Learning Methods to compress signals from various channels, (ii) builds a unique representation of each trial by concatenating the channels' representations into a unique low-dimensional vector and (iii) selects from these vectors the most relevant channels to perform classification. After training, the algorithm can be exploited by (iv) transferring the parametrized subgroup of selected channel-specific RLMs to new signals and (v) obtain novel trial vectors to be fed to any kind of classifier. We tested the algorithm on a case study attaining extremely promising results when considering the complexity of subject independent channel selection. |
-
61/2020 - 10/09/2020
Pozzi, S.; Redaelli, A.; Vergara, C.; Votta, E.; Zunino, P.
Mathematical and numerical modeling of atherosclerotic plaque progression based on fluid-structure interaction | Abstract | | In this work we propose a mathematical and numerical model to describe the early stages of atherosclerotic plaque formation,
which is based on the interaction of processes with different spatial and temporal scales.
A fluid-structure interaction problem, used to describe the cardiovascular mechanics arising between blood and the artery wall, is coupled to a set of differential problems describing the evolution of solute concentrations.
In order to manage the multiscale-in-space nature of the involved processes, we propose a suitable numerical strategy based on the splitting and sequential solution of the coupled problem.
We present some preliminary numerical results and investigate the effects of geometry, model parameters and coupling strategy on plaque growth. |
-
60/2020 - 19/08/2020
Lupo Pasini, M; Perotto, S.
Hierarchical model reduction driven by a Proper Orthogonal Decomposition for parametrized advection-diffusion-reaction problems | Abstract | | This work combines the Hierarchical Model (HiMod) reduction technique with a standard Proper Orthogonal Decomposition (POD) to solve parametrized partial differential equations modeling advection-diffusion-reaction phenomena in elongated domains (e.g., pipes). This combination leads to what we define a HiPOD model reduction, which merges the reliability of HiMod with the computational efficiency of POD. Two different HiPOD techniques are presented and assessed through an extensive numerical verification. |
-
59/2020 - 07/08/2020
Massi, M.C.; Franco, N.R; Ieva, F.; Manzoni, A.; Paganoni, A.M.; Zunino, P.
High-Order Interaction Learning via Targeted Pattern Search | Abstract | | Logistic Regression (LR) is a widely used statistical method in empirical studies in many research fields. However, these real-life scenarios oftentimes share complexities that would hinder the application of the as-is model. First and foremost, the need to include high-order interactions to capture the variability of their data. Moreover, these studies are seldom developed in imbalanced settings, with datasets growing wider, sample size
from very large to extremely small and a strong need for model and results interpretability.
In this paper we present a novel algorithm, High-Order Interaction Learning via targeted Pattern search (HOILP), to select interaction terms of varying order to include in a LR for
an imbalanced binary classification task when input data is categorical. HOILP’s rationale is built on the duality between item sets and categorical interactions, and is composed of
(i) an interaction learning step based on a well-known frequent item set mining algorithm and (ii) a novel dissimilarity-based interaction selection step, that allows the user to control
for the number of interactions to include in the LR model. Besides HOILP we present here two variants (Scores HOILP and Clusters HOILP), that can suit even more specific needs.
Through a set of experiments we validate our algorithm and prove its wide applicability to real-life research scenarios, surpassing the performance of a benchmark state-of-the-art
algorithm. |
-
58/2020 - 07/08/2020
Beraha, M.; Pegoraro, M.; Peli, R.; Guglielmi, A
Spatially dependent mixture models via the Logistic Multivariate CAR prior | Abstract | | We consider the problem of spatially dependent areal data, where for each
area independent observations are available, and propose to model
the density of each area through a finite mixture of Gaussian distributions.
The spatial dependence is introduced via a novel joint distribution for
a collection of vectors in the simplex, that we term logisticMCAR.
We show that salient features of the logisticMCAR distribution
can be described analytically, and that a suitable augmentation scheme based on the
P{'o}lya-Gamma identity allows to derive an efficient Markov Chain Monte Carlo
algorithm.
When compared to competitors, our model has proved to better estimate densities in different (disconnected) areal locations when they have different characteristics.
We discuss an application on a real dataset of Airbnb listings in the city
of Amsterdam, also showing how to easily incorporate for additional covariate
information in the model. |
|
