191 research outputs found
Presentazione
No full textBreve introduzione al volume per l’anniversario di fondazione della scuola di arti e mestieri Ricchino di Rovato con riguardo alle linee educative.Brief introduction to the volume on the anniversary of the founding of the school of arts and crafts, Francesco Ricchino, of Rovato regarding to educational lines
Con lo sguardo della storia. Premessa,
No full textSaggio introduttivo al volume relativo ai percorsi formativi, attivati a Rovato (Brescia), sul lavoro dello storico, il metodo scientifico di indagine e la ricerca d’archivio; i percorsi di apprendimento, insieme ad alcuni degli esiti di studio, corredano il volume
"Obsculta praecepta magistri". Monaci e monasteri medievali in terra bresciana
No full textSi illustra la diffusione del monachesimo in ambito bresciano dall'età longobarda alla fine del medioevo; alla presentazione dei singoli cenobi segue il quadro della fondazione dei priorati cluniacensi
Heterogeneous Datasets for Federated Survival Analysis Simulation
No full textHeterogeneous Datasets for Federated Survival Analysis Simulation
This repo contains three algorithms for constructing realistic federated datasets for survival analysis. Each algorithm starts from an existing non-federated dataset and assigns each sample to a specific client in the federation. The algorithms are:
uniform_split: assigns each sample to a random client with uniform probability;
quantity_skewed_split: assigns each sample to a random client according to the Dirichlet distribution [3, 4];
label_skewed_split: assigns each sample to a time bin, then assigns a set of samples from each bin to the clients according to the Dirichlet distribution [3, 4].
For more information, please take a look at our paper at https://arxiv.org/abs/2301.12166 [1].
Content
federated_survival_datasets.zip: the content of the repository at https://github.com/archettialberto/federated_survival_datasets
Heterogheneous_Datasets_for_Federated_Survival_Analysis_Simulation.pdf: the conference paper describing the work.
Installation
Federated Survival Datasets is built on top of numpy and scikit-learn. To install those libraries you can run pip install -r requirements.txt. To import survival datasets into your project, we strongly recommend SurvSet (https://github.com/ErikinBC/SurvSet) [2], a comprehensive collection of more than 70 survival datasets.
Usage
import numpy as np
import pandas as pd
from federated_survival_datasets import label_skewed_split
# import a survival dataset and extract the input array X and the output array y
df = pd.read_csv("metabric.csv")
X = df[[f"x{i}" for i in range(9)]].to_numpy()
y = np.array([(e, t) for e, t in zip(df["event"], df["time"])], dtype=[("event", bool), ("time", float)])
# run the splitting algorithm
client_data = label_skewed_split(num_clients=8, X=X, y=y)
# check the number of samples assigned to each client
for i, (X_c, y_c) in enumerate(client_data):
print(f"Client {i} - X: {X_c.shape}, y: {y_c.shape}")
We provide an example notebook in the zipped folder to illustrate the proposed algorithms. It requires scikit-survival, seaborn, and pandas.
References
[1] Archetti, A., Lomurno, E., Lattari, F., Martin, A., & Matteucci, M. (2023). Heterogeneous Datasets for Federated Survival Analysis Simulation. arXiv preprint arXiv:2301.12166.
[2] Drysdale, E. (2022). SurvSet: An open-source time-to-event dataset repository. arXiv preprint arXiv:2203.03094.
[3] Hsu, T. M. H., Qi, H., & Brown, M. (2019). Measuring the effects of non-identical data distribution for federated visual classification. arXiv preprint arXiv:1909.06335.
[4] Li, Q., Diao, Y., Chen, Q., & He, B. (2022, May). Federated learning on non-iid data silos: An experimental study. In 2022 IEEE 38th International Conference on Data Engineering (ICDE) (pp. 965-978). IEEE
Schede bibliografiche, nr. 268, 315, 322, 325, 338, 339, 345, 357, 359, 369, 371, 377, 381, 496
No full textQuattordici schede bibliografiche relative alla storia della Chiesa in Italia, dal medioevo al Novecento, relative alle aree lombarda e trentina; tra i temi trattati si segnalano i sermoni di Alessio da Seregno, la tradizione eucaristica nella liturgia ambrosiana, le chiese di Asola, di San Zavedro e di Mezzolombardo, la formazione musicale in S. Francesco a Brescia, le scuole parrocchiali e la formazione del clero in Valcamonica di età moderna, il canonico bresciano Paolo Gagliardi, l’istituzione delle dimesse a Chiari tra XVII e XIX secolo, la figura di Giovanni Guadagnini e il giansenismo, il funzionamento della Congrega della carità apostolica di Brescia, il clero diocesano e la prima guerra mondiale nelle carte dell’Archivio storico diocesano di Brescia, il rinnovamento artistico della Chiesa di San Giovanni a Brescia e il recupero dell’arte sacra medievale nella formazione dell’identità nazionale Italia tra Otto e Novecento. Pagine complessive: 12.Fourteen bibliographic records have been drawn up relating to the history of the Church in Italy, from the Middle Ages to the twentieth century, relating to the Lombard and Trentino areas. Among the topics covered are the sermons of Alessio da Seregno, the Eucharistic tradition in the Ambrosian liturgy, the churches of Asola, of San Zavedro and of Mezzolombardo, the musical training in S. Francesco in Brescia, the parochial schools and the formation of the clergy in modern Valcamonica, the Brescia canon Paolo Gagliardi, the institution of the nuns resigned in Chiari between the seventeenth and nineteenth centuries, the figure of Giovanni Guadagnini and Jansenism, the functioning of the Congregation of the apostolic charity of Brescia, the diocesan clergy and the First World War in the documentation of the Diocesan Historical Archive of Brescia, the artistic renewal of the Church of San Giovanni in Brescia and the recovery of medieval sacred art in the formation of the national identity of Italy between the nineteenth and twentieth centuries. Total pages: 12
Salute fisica e spirituale. Note sparse dal monachesimo altomedievale
No full textI temi del benessere e della felicità appartengono a ogni società. Per il medioevo una chiave di lettura è data dalla cultura monastica, dove la salute del corpo e quella dell’anima sono strettamente correlate. Le scelte alimentari sono fondamentali nell’orientare i regimi dietetici che sono suggeriti da testi e regole monastiche con grande varietà di indicazioni. Se non esistono cibi proibiti, non tutti favoriscono il percorso spirituale del monaco, la cui ascesi passa in primo luogo – ma non solo – attraverso un regime alimentare disciplinato. I presupposti biblici e spirituali alla base degli usi alimentari, tuttavia, acquistano piena luce con le conoscenze medico-dietetiche, grazie a cui si stabilivano i cibi da mangiare, con quali abbinamenti e l’ora dei pasti. Qualche mirata annotazione viene dedicata al contesto iberico.The themes of well-being and happiness belong to every society. For the Middle Ages a key to understanding this is provided by monastic culture, where the health of the body and that of the soul are closely related. Food choices are fundamental in guiding dietary regimes that are suggested by monastic texts and rules with a great variety of indications. The biblical and spiritual assumptions underlying dietary customs, however, gain full light with the medico-dietetic knowledge, thanks to which the foods to be eaten, with which pairings and the time of meals were determined. A few targeted annotations are devoted to the Iberian context
The Set Orienteering Problem
No full textIn this paper, we study the Set Orienteering Problem which is a generalization of the Orienteering Problem where customers are grouped in clusters and a profit is associated with each cluster. The profit of a cluster is collected only if at least one customer from the cluster is visited. A single vehicle is available to collect the profit and the objective is to find the vehicle route that maximizes the profit collected and such that the route duration does not exceed a given threshold. We propose a mathematical formulation of the problem and a matheuristic algorithm. Computational tests are made on instances derived from benchmark instances for the Generalized Traveling Salesman Problem with up 1084 vertices. Results show that the matheuristic produces robust and high-quality solutions in a short computing time
"Servizio buono e commendevole". Brixia sacra: cento anni di storia della Chiesa
No full textSorta nel 1910 per l'impegno di Paolo Guerrini la rivista "Brixia Sacra" viene indagata nelle sue finalità e nelle sue trasformazioni nel corso di un secolo di vita; attenta alla storia della Chiesa bresciana, svolse un ruolo centrale nell'avvio della Rivista di storia della Chiesa in Italia ed ebbe importanti e constanti rapporti con uomini della cultura e delle istituzioni ecclesiastiche del Novecento (Placido Lugano, Pio Paschini, Giovanni Battista Montini, Angelo Roncalli, Michele Maccarrone, Giuseppe De Luca, Nello Vian, Francesco Traniello, ecc.).The essay studies the purposes and the transformations over the course of the first one century of life of the review “Brixia Sacra", born in 1910 for the strong will of the priest Paolo Guerrini. The review was careful to the Church history in Brescia and it played a central role in Italy at the birth of the first Reviews of Church history: it also had important relationships with men of culture and ecclesiastical institutions of the twentieth century (like Placido Lugano, Pius Paschini, Giovanni Battista Montini, Angelo Roncalli, Michele Maccarrone, Giuseppe De Luca, Nello Vian, Francesco Traniello, etc.)
Bridging the gap: improve neural survival models with interpolation techniques
No full textSurvival analysis is an essential tool in healthcare for risk assessment, assisting clinicians in their evaluation and decision making processes. Therefore, the importance of using expressive and high-performing survival models is crucial. With the advent of neural networks and deep learning, a new generation of survival models has emerged, offering state-of-the-art capabilities to capture the non-linear and complex relationships inherent in multimodal patient data for survival prediction. However, these models often produce discrete outputs, resulting in survival functions that are coarse-grained and difficult to interpret. This study advances previous research by further exploring interpolation techniques as a post-processing strategy to improve the predictive accuracy of survival models. Our results show how the use of specific interpolation techniques significantly improves the concordance and calibration of survival estimates. This analysis encompasses a wide array of medical datasets, models, and interpolation techniques, demonstrating the effectiveness of the proposed approach and providing actionable insights for survival model design
Deep Survival Analysis for Healthcare: An Empirical Study on Post-Processing Techniques
No full textSurvival analysis is a crucial tool in healthcare, allowing us to understand and predict time-to-event occurrences using statistical and machine-learning techniques. As deep learning gains traction in this domain, a specific challenge emerges: neural network-based survival models often produce discrete-time outputs, with the number of discretization points being much fewer than the unique time points in the dataset, leading to potentially inaccurate survival functions. To this end, our study explores post-processing techniques for survival functions. Specifically, interpolation and smoothing can act as effective regularization, enhancing performance metrics integrated over time, such as the Integrated Brier Score and the Cumulative Area-Under-the-Curve. We employed various regularization techniques on diverse real-world healthcare datasets to validate this claim. Empirical results suggest a significant performance improvement when using these post-processing techniques, underscoring their potential as a robust enhancement for neural network-based survival models. These findings suggest that integrating the strengths of neural networks with the non-discrete nature of survival tasks can yield more accurate and reliable survival predictions in clinical scenarios
- …
