International Scientific Journal for Research

Data mining uses machine learning algorithms

PAWAN WHIG — Sun, 08 Dec 2019 00:00:00 +0000

We use learning algorithms every day in many different applications. One of the reasons an internet search engine like Google or Bing works so effectively whenever it is used to look the web is that a learning formula, one used by Google or Microsoft, has really discovered how to grade websites. Artificial intelligence is also used every time Facebook is used and also it recognises friends' images. Several machine learning methods have really been reviewed in this study. These algorithms are used for a variety of tasks including data mining, image processing, predictive analytics, and more.

Machine learning validation guidelines in biology

Radey Krishan — Sun, 17 Nov 2019 00:00:00 +0000

Machine learning is often used in modern biology to give predictions and enhance decision-making processes. There has recently been a request for more investigation of machine learning performance and its limitations. We provide a collection of community-wide suggestions aimed at establishing machine learning validation standards in biology. Implementing a structured methods definition for machine learning based on DOME (data, optimization, model, evaluation) can help both reviewers and readers better understand and analyse a method's or outcome's performance and limits.

Forward with a statistical model for modelling Neural Networks

Ankit Sharma — Mon, 19 Aug 2019 00:00:00 +0000

Even though neural networks are extensively employed in a wide range of applications, they are still regarded as black boxes and pose certain challenges in terms of dimensioning and measuring their prediction inaccuracy. This has sparked increased interest in the overlapping region between neural networks and more classic statistical approaches, which can aid in overcoming such issues. In this article, a mathematical framework linking neural networks and polynomial regression is investigated by employing a Taylor expansion technique to construct an explicit expression for the coefficients of a polynomial regression from the weights of a particular neural network. This is accomplished in regression issues using single hidden layer neural networks.

A Machine Learning Algorithm to Identify Patients

Raman Verma — Sun, 15 Dec 2019 00:00:00 +0000

Risk categorization of individual infection-prone individuals would enable surgeons to closely monitor high-risk patients and take appropriate action early on. This can lessen the negative effects of infection, such higher medical expenses. In order to determine the likelihood of infection in patients with surgically treated tibial shaft fractures, this study developed a machine learning (ML)-derived risk-stratification tool using the SPRINT .

Enhancing Classification Precision: Exploring the Power of Support-Vector Networks in Machine Learning

Balaram Yadav Kasula — Sat, 17 Aug 2019 00:00:00 +0000

Support-Vector Networks (SVNs) have emerged as powerful tools in the realm of machine learning, offering robust classification capabilities and efficient handling of high-dimensional data. This paper presents an in-depth exploration of the principles, applications, and advancements in support-vector networks within the context of machine learning paradigms. The abstract nature of SVNs, encapsulating a kernel-based approach for pattern recognition and classification, underscores their adaptability to complex datasets, rendering them invaluable in various domains. Key aspects covered include the foundational principles of SVNs, their optimization techniques, and their applicability in diverse scenarios, such as image recognition, natural language processing, and bioinformatics. Moreover, the paper delves into the comparative analysis of SVNs with other classification algorithms, highlighting their strengths and limitations. Furthermore, considerations regarding parameter tuning, scalability, and interpretability are discussed. This comprehensive review aims to offer insights into the multifaceted utility of support-vector networks, underlining their significance as a cornerstone in the machine learning landscape.

AI-Enhanced Drug Discovery: Accelerating the Development of Targeted Therapies

Venkata Sai Teja Yarlagadda — Sat, 17 Aug 2019 00:00:00 +0000

The process of drug discovery is complex, time-consuming, and expensive. This paper explores how AI and machine learning are being used to accelerate drug discovery by predicting the efficacy of potential drug compounds, identifying new drug targets, and optimizing clinical trial designs. We examine the application of deep learning algorithms, such as neural networks and reinforcement learning, in screening large chemical libraries and predicting drug interactions. The paper also discusses the integration of AI in personalized medicine, where AI can help identify patients who would benefit most from specific therapies, thus reducing trial and error in drug development.