Questioning the reliability of Monte Carlo simulation for machine learning test validation

2016 ◽  
Author(s):  
Gildas Leger ◽  
Manuel J. Barragan
Keyword(s):  
Machine Learning ◽  
Monte Carlo Simulation ◽  
Monte Carlo ◽  
Test Validation ◽  
Learning Test

scholarly journals 3D convolutional GAN for fast simulation

2019 ◽  
Vol 214 ◽  
pp. 02010 ◽  
Author(s):  
Sofia Vallecorsa ◽  
Federico Carminati ◽  
Gulrukh Khattak
Keyword(s):  
Machine Learning ◽  
Monte Carlo Simulation ◽  
Monte Carlo ◽  
Three Dimensional ◽  
Machine Learning Techniques ◽  
Generative Adversarial Networks ◽  
Detector Response ◽  
Fast Simulation ◽  
Geant4 Monte Carlo Simulation ◽  
Detector Simulation

Machine Learning techniques have been used in different applications by the HEP community: in this talk, we discuss the case of detector simulation. The need for simulated events, expected in the future for LHC experiments and their High Luminosity upgrades, is increasing dramatically and requires new fast simulation solutions. We describe an R&D activity aimed at providing a configurable tool capable of training a neural network to reproduce the detector response and speed-up standard Monte Carlo simulation. This represents a generic approach in the sense that such a network could be designed and trained to simulate any kind of detector and, eventually, the whole data processing chain in order to get, directly in one step, the final reconstructed quantities, in just a small fraction of time. We present the first application of three-dimensional convolutional Generative Adversarial Networks to the simulation of high granularity electromagnetic calorimeters. We describe detailed validation studies comparing our results to Geant4 Monte Carlo simulation. Finally we show how this tool could be generalized to describe a whole class of calorimeters, opening the way to a generic machine learning based fast simulation approach.

Simulate, Optimise, Partition: Algorithmic Diagrams of Pattern Recognition from 1953 Onwards

2016 ◽  
Author(s):  
Adrian Mackenzie
Keyword(s):  
Machine Learning ◽  
Data Mining ◽  
Pattern Recognition ◽  
Monte Carlo Simulation ◽  
Monte Carlo ◽  
Deep Learning ◽  
Clustering Algorithms ◽  
Flow Through ◽  
The Cold War ◽  
Hand Writing

Contemporary attempts to find patterns in data, ranging from the now mundane technologies of hand-writing recognition through to mammoth infrastructure-heavy practices of deep learning conducted by major business and government actors, rely on a group of techniques intensively developed during the 1950-60s in physics, engineering and psychology. Whether we designate them as pattern recognition, data mining, or machine learning, these techniques all seek to uncover patterns in data that cannot appear directly to the human eye, either because there are too many items for anyone to look at, or because the patterns are too subtly woven through in the data. From the techniques in current use, three developed in the Cold War era iconify contemporary modes of pattern finding: Monte Carlo simulation, gradient descent, and clustering algorithms that search for groups or clusters in data. Each of these techniques implements a different mode of pattern, and these different modes of pattern recognition flow through into contemporary scientific, technological, business and governmental problematizations. The different perspectives on event, trajectory, and proximity they embody imbue many power relations, forms of value and the play of truth/falsehood today.

A Guide to Monte Carlo Simulations in Statistical Physics

2021 ◽  
Author(s):  
David Landau ◽  
Kurt Binder
Keyword(s):  
Machine Learning ◽  
Monte Carlo Simulation ◽  
Monte Carlo ◽  
Graduate Students ◽  
Physical Science ◽  
Statistical Physics ◽  
Condensed Matter ◽  
Physical Systems ◽  
Recent Developments ◽  
New Material

Dealing with all aspects of Monte Carlo simulation of complex physical systems encountered in condensed matter physics and statistical mechanics, this book provides an introduction to computer simulations in physics. The 5th edition contains extensive new material describing numerous powerful algorithms and methods that represent recent developments in the field. New topics such as active matter and machine learning are also introduced. Throughout, there are many applications, examples, recipes, case studies, and exercises to help the reader fully comprehend the material. This book is ideal for graduate students and researchers, both in academia and industry, who want to learn techniques that have become a third tool of physical science, complementing experiment and analytical theory.

scholarly journals Machine Learning & Mechanics of “Investment Matrix”: “Performance Optimisation & Risk Measurement of Bank Nifty”

2020 ◽  
Vol 8 (6) ◽  
pp. 3298-3302
Keyword(s):  
Machine Learning ◽  
Monte Carlo Simulation ◽  
Monte Carlo ◽  
Risk Analysis ◽  
Practical Implication ◽  
Moving Average ◽  
Technical Analysis ◽  
Research Paper ◽  
Risk Measurement ◽  
Analysis Tool

Purpose: The zeal and reason to write this research paper are to evaluate the performance & risk measurement of Bank Nifty based on Machine learning, Technical Analysis & Monte Carlo Simulation. Design /Methodologies/Approach: To achieve our desired results for this study, we use moving average (auto-optimization method) as a technical analysis return optimization tool & Monte Carlo Simulation as a risk analysis tool, & at the end harmonize both of the results, & compare with buy hold strategy. We use Bank Nifty end of day historical closing data of past five years i.e.1 Jan 2015 – 31 Dec 2019, For this study using Amibroker software. Originality & Value: This research paper is beneficial for anyone who wants understand Bank Nifty on the ground of technical analysis & risk measurement technique (MCs), & also to synergies the strength of two studies. Research Limitations: In appropriate input can lead to creating wrong simulation result, there are no. of unknown factors that simulation cannot truly understand or account during the process. Practical implication: Understanding stock market results is essential to make further decisions related to risk & reward ratio. The results imply that Moving average give outstanding returns on Bank Nifty in medium to long run, & Monte Carlo Simulation having functional judgemental abilities on probabilities basis towards risk & returns. Furthermore, by apply both the technique for risk analysis, simultaneously give outstanding risk & return optimization of Bank Nifty.

Sign in / Sign up

Export Citation Format

Share Document