scholarly journals Reaction-Based Machine Learning Representations for Predicting the Enantioselectivity of Organocatalysts

2021 ◽  
Author(s):  
Simone Gallarati ◽  
Raimon Fabregat ◽  
Ruben Laplaza ◽  
Sinjini Bhattacharjee ◽  
Matthew D. Wodrich ◽  
...  
Keyword(s):  
Machine Learning ◽  
High Purity ◽  
Computational Design ◽  
Chiral Compounds ◽  
Significant Challenge

HHundreds of catalytic methods are developed each year to meet the demand for high-purity chiral compounds. The computational design of enantioselective organocatalysts remains a significant challenge, as catalysts are typically...

scholarly journals A Sound-based Machine Learning to Predict Traffic Vehicle Density

2021 ◽  
Vol 9 (1) ◽  
pp. 55-62
Author(s):  
Geoferleen Flores ◽  
◽  
Eduardo Jr. Piedad ◽  
Anzeneth Figueroa ◽  
Romari Tumamak ◽  
...  
Keyword(s):  
Machine Learning ◽  
Traffic Flow ◽  
Data Gathering ◽  
Sound Intensity ◽  
Support Vector ◽  
Mean Square ◽  
Significant Challenge ◽  
Vehicle Density ◽  
Smart Technologies ◽  
Machine Learning Models

Traffic flow mismanagement is a significant challenge in all countries especially in crowded cities. An alternative solution is to utilize smart technologies to predict traffic flow. In this study, frequency spectrum describing traffic sound characteristics is used as an indicator to predict the next five-minute vehicle density. Sound frequency and vehicle intensity are collected during a thirteen-hour data gathering. The collected sound intensity and frequency are then used to learn three machine-learning models - support vector machine, artificial neural network, and random forest and to predict vehicle intensity. It was found out that the performances of the three models based on root-mean-square-error values are 12.97, 16.01, and 10.67, respectively. These initial and satisfactory results pave a new way to predict traffic flow based on traffic sound characteristics which may serve as a better alternative to conventional features.

scholarly journals A Novel Stacked Ensemble for Hate Speech Recognition

2021 ◽  
Vol 11 (24) ◽  
pp. 11684
Author(s):  
Mona Khalifa A. Aljero ◽  
Nazife Dimililer
Keyword(s):  
Machine Learning ◽  
Hate Speech ◽  
State Of The Art ◽  
Majority Voting ◽  
Support Vector ◽  
Timely Manner ◽  
Machine Learning Classifiers ◽  
Significant Challenge ◽  
Final Output ◽  
Harmful Content

Detecting harmful content or hate speech on social media is a significant challenge due to the high throughput and large volume of content production on these platforms. Identifying hate speech in a timely manner is crucial in preventing its dissemination. We propose a novel stacked ensemble approach for detecting hate speech in English tweets. The proposed architecture employs an ensemble of three classifiers, namely support vector machine (SVM), logistic regression (LR), and XGBoost classifier (XGB), trained using word2vec and universal encoding features. The meta classifier, LR, combines the outputs of the three base classifiers and the features employed by the base classifiers to produce the final output. It is shown that the proposed architecture improves the performance of the widely used single classifiers as well as the standard stacking and classifier ensemble using majority voting. We also present results on the use of various combinations of machine learning classifiers as base classifiers. The experimental results from the proposed architecture indicated an improvement in the performance on all four datasets compared with the standard stacking, base classifiers, and majority voting. Furthermore, on three of these datasets, the proposed architecture outperformed all state-of-the-art systems.

scholarly journals Linking synthesis and structure descriptors from a large collection of synthetic records of zeolite materials

2019 ◽  
Vol 10 (1) ◽  
Author(s):  
Koki Muraoka ◽  
Yuki Sada ◽  
Daiki Miyazaki ◽  
Watcharop Chaikittisilp ◽  
Tatsuya Okubo
Keyword(s):  
Machine Learning ◽  
Machine Learning Techniques ◽  
Empirical Knowledge ◽  
Large Collection ◽  
Synthesis Conditions ◽  
Significant Challenge ◽  
Learning Techniques ◽  
Structure Relationship ◽  
Similar Materials ◽  
Crossover Experiments

Abstract Correlating synthesis conditions and their consequences is a significant challenge, particularly for materials formed as metastable phases via kinetically controlled pathways, such as zeolites, owing to a lack of descriptors that effectively illustrate the synthesis protocols and their corresponding results. This study analyzes the synthetic records of zeolites compiled from the literature using machine learning techniques to rationalize physicochemical, structural, and heuristic insights to their chemistry. The synthesis descriptors extracted from the machine learning models are used to identify structure descriptors with the appropriate importance. A similarity network of crystal structures based on the structure descriptors shows the formation of communities populated by synthetically similar materials, including those outside the dataset. Crossover experiments based on previously overlooked structural similarities reveal the synthesis similarity of zeolites, confirming the synthesis–structure relationship. This approach is applicable to any system to rationalize empirical knowledge, populate synthesis records, and discover novel materials.

Malicious URL Detection Using Machine Learning

2021 ◽  
pp. 160-180
Author(s):  
Ferhat Ozgur Catak ◽  
Kevser Sahinbas ◽  
Volkan Dörtkardeş
Keyword(s):  
Machine Learning ◽  
Random Forest ◽  
Web Sites ◽  
Internet Usage ◽  
Gradient Boosting ◽  
Malicious Software ◽  
Significant Challenge ◽  
Forest Models ◽  
Random Forest Models ◽  
Security Components

Recently, with the increase in Internet usage, cybersecurity has been a significant challenge for computer systems. Different malicious URLs emit different malicious software and try to capture user information. Signature-based approaches have often been used to detect such websites and detected malicious URLs have been attempted to restrict access by using various security components. This chapter proposes using host-based and lexical features of the associated URLs to better improve the performance of classifiers for detecting malicious web sites. Random forest models and gradient boosting classifier are applied to create a URL classifier using URL string attributes as features. The highest accuracy was achieved by random forest as 98.6%. The results show that being able to identify malicious websites based on URL alone and classify them as spam URLs without relying on page content will result in significant resource savings as well as safe browsing experience for the user.

scholarly journals Towards computational design of zeolite catalysts for CO2 reduction

RSC Advances ◽  
2015 ◽  
Vol 5 (55) ◽  
pp. 44361-44370 ◽  
Author(s):  
A. W. Thornton ◽  
D. A. Winkler ◽  
M. S. Liu ◽  
M. Haranczyk ◽  
D. F. Kennedy
Keyword(s):  
Machine Learning ◽  
Molecular Simulation ◽  
Co2 Reduction ◽  
Computational Design ◽  
Zeolite Catalysts ◽  
Structure Database ◽  
Computational Search ◽  
Reduction Catalysts

Computational search of structure database for CO2 reduction catalysts using molecular simulation and machine learning.

scholarly journals Machine Learning for Fabrication of Graded Knitted Membranes

2021 ◽  
pp. 309-319
Author(s):  
Yuliya Sinke ◽  
Sebastian Gatz ◽  
Martin Tamke ◽  
Mette Ramsgaard Thomsen
Keyword(s):  
Machine Learning ◽  
Integrated Design ◽  
Computational Design ◽  
Parametric Models ◽  
Performance Criteria ◽  
Data Generation ◽  
Research Projects ◽  
Training Examples ◽  
Set Up

AbstractThis paper examines the use of machine learning in creating digitally integrated design-to-fabrication workflows. As computational design allows for new methods of material specification and fabrication, it enables direct functional grading of material at high detail thereby tuning the design performance in response to performance criteria. However, the generation of fabrication data is often cumbersome and relies on in-depth knowledge of the fabrication processes. Parametric models that set up for automatic detailing of incremental changes, unfortunately, do not accommodate the larger topological changes to the material set up. The paper presents the speculative case study KnitVault. Based on earlier research projects Isoropia and Ombre, the study examines the use of machine learning to train models for fabrication data generation in response to desired performance criteria. KnitVault demonstrates and validates methods for shortcutting parametric interfacing and explores how the trained model can be employed in design cases that exceed the topology of the training examples.

scholarly journals Log Message Anomaly Detection with Oversampling

2020 ◽  
Author(s):  
Amir Farzad ◽  
T. Aaron Gulliver
Keyword(s):  
Machine Learning ◽  
Feature Extraction ◽  
Anomaly Detection ◽  
Learning Algorithms ◽  
Imbalanced Data ◽  
Machine Learning Algorithms ◽  
Detection Accuracy ◽  
Data Sets ◽  
Significant Challenge ◽  
Proposed Model

Imbalanced data is a significant challenge in classification with machine learning algorithms. This is particularly important with log message data as negative logs are sparse so this data is typically imbalanced. In this paper, a model to generate text log messages is proposed which employs a SeqGAN network. An Autoencoder is used for feature extraction and anomaly detection is done using a GRU network. The proposed model is evaluated with three imbalanced log data sets, namely BGL, OpenStack, and Thunderbird. Results are presented which show that appropriate oversampling and data balancing improves anomaly detection accuracy.

scholarly journals Machine learning and computational design

Ubiquity ◽  
2020 ◽  
Vol 2020 (May) ◽  
pp. 1-10
Author(s):  
Silvio Carta
Keyword(s):  
Machine Learning ◽  
Computational Design

scholarly journals One-Step Synthesis and Characterization of Silica Nano-/Submicron Spheres by Catalyst-Assisted Pyrolysis of a Preceramic Polymer

2013 ◽  
Vol 2013 ◽  
pp. 1-4
Author(s):  
Feng Gao ◽  
Zhijian Peng ◽  
Xiuli Fu
Keyword(s):  
High Purity ◽  
Large Scale ◽  
Synthesis And Characterization ◽  
Silica Nanospheres ◽  
Preceramic Polymer ◽  
Simple Method ◽  
Significant Challenge ◽  
One Step ◽  
Controlled Morphology

Silica nanospheres have attracted tremendous interest due to their importance in extensive applications. However, the direct large-scale fabrication of silica nanospheres with controlled morphology and high purity remains a significant challenge. In this work, silica nano-/submicron spheres were successfully synthesized by a simple method through pyrolysis of an amorphous polysilazane preceramic powder with catalyst FeCl2. The synthesized spheres possess well-designed shape with diameter of 600–800 nm and high purity. The surfaces of the spheres are smooth and clean without any flaws. Besides, the spheres are identified as amorphous silica, and their growth mechanism was also proposed.

Sign in / Sign up

Export Citation Format

Share Document