A Multi-Source Information Retrieval System for Question Answering with Sentiment Analysis and LLM Interpretation

Published: March 14, 2024

In the era of ubiquitous digital information, question-answering systems have become indispensable tools for accessing and extracting knowledge from vast datasets. This research explores design, implementation, evaluation a system capable processing user queries, retrieving information diverse sources, generating accurate responses. The study focuses on assessing performance under three operational modes: offline alone, online combined (offline online). A meticulously curated dataset comprising queries corresponding ground truth answers forms basis experimental evaluation. Performance metrics including response time, satisfaction, coverage are evaluated to gain insights into functionality efficacy across different paradigms. findings highlight significance leveraging both data sources optimize performance, underscoring importance balanced approach in designing intelligent systems. contributes advancing state-of-the-art technology lays foundation future developments aimed at enhancing retrieval extraction capabilities various domains.

Language: Английский

---

Machine Learning Model for Forecasting the Rating of Mobile Apps

Published: March 14, 2024

Customers' decisions are greatly influenced by what they read or see online. Reviews customers demonstrate their knowledge about quality and experience. In the Google Play store, applications' success can be significantly phoney numerical ratings. It's well that a positive review may strongly associated with high star rating. Nevertheless, text format of reviews typically differs from information provided user The effective machine learning approach for forecasting app ratings on Store is displayed in this study.

Language: Английский

---

An efficient swarm intelligence algorithm for Multi-Objective Task Scheduling Optimization in the Context of Cloud Computing

Published: March 14, 2024

In the modern era cloud computing is a fastest growing field. It provides many facilities to user store, manipulate and retrieve data anytime anywhere with help of web or internet. Time completion tasks load balancing virtual machines are two significant issues that need fixing in settings. This study suggests Multi-objective Optimisation technique overcome these issues. GA, PSO ACO Algorithm utilized schedule multiple objectives. Makespan, critical parameter measuring overall time necessary accomplish all jobs, primary focus since it assesses performance various algorithms great detail. our research, produces makespan 1.1988, GA 1.3025, while remarkably achieves minimum 0.8736 after 200 iterations. The comparison's findings demonstrate better making ideal choice for task scheduling algorithm optimization. difference values demonstrates ACO's superior capacity explore solution space, converge effectively, offer plans minimize amount required task. this crucial because they advantages several objectives efficiently at once.

Language: Английский

---

Detection Of Breast Tumour Using Advanced Methods

Published: March 15, 2024

Breast cancer is a dangerous disease that can be fatal if left unchecked. It the main reason people die in females and occurs both men women. Adenocarcinoma most prevalent kind of breast cancer, early diagnosis crucial for effective treatment. A hybrid model combining several machine learning algorithms has been proposed to diagnose adenocarcinoma accurately. The based on ensemble easy implement using cheaper safer methods. Random Forest provides maximum accuracy (86-90%) with least amount error.

Language: Английский

---

Classification of Sign Language Gestures using CNN with Adam Optimizer

Published: March 15, 2024

In this paper, we have worked on to improve the overall accuracy and individual accuracies of given gestures before developing sign language prediction system for that taken a dataset pre-processed images from Kaggle further applied CNN model with four layers relu activation function, has three pooling layer softmax after execution got 99.95%.

Language: Английский

---

Quantitative Insights into Breakout and Reversal Trading Strategies: A Comprehensive Analysis for Algorithmic Trading in Financial Markets

Published: March 15, 2024

This study explores trading strategies focused on breakout and reversal techniques providing insights, for participants in today's markets. Rooted the realms of quantitative finance research delves into two approaches that aim to identify moments swing trading. The analysis places emphasis strategy, which combines closing price candlesticks with highest observed over previous 20 candlesticks. By taking account volume this strategy becomes a reliable indicator identifying stocks are likely experience substantial directional movements. chosen time frame effectively captures shifts while minimizing any noise intervals. On hand focuses spotting divergences within Relative Strength Index (RSI). Recognizing RSI has ability signal changes momentum approach identifies undergoing change direction contrary their Specific criteria include an reading followed by low aligning lows stock's price. To validate these empirically historical stock market data is used back-testing purposes. findings provide evaluation performance, across-market conditions offering valuable insights effectiveness potential real-world applications. aims assist traders investors changing world finance. It provides method, navigating complexities markets explaining strategies, including performance. intends empower knowledge helping them make decisions implement effective strategies.

Language: Английский

---

An Ensemble Deep Learning Framework for Binary Classification of COVID-19 Cases Using the VGG16 (Visual Geometry Group 16) Architecture: A Comprehensive Analysis and Performance Evaluation

Published: March 15, 2024

This paper presents a novel approach to deep learning by putting forth cooperative system that uses the VGG16 architecture categorise COVID-19 examples into two groups. Our model is distinguished its remarkable recall metrics and precision, which achieve careful balance essential for accurate categorization. What's more impressive how well performs in non-COVID category, effectively differentiating between COVID cases. With overall accuracy of 96%, successfully classifies cases from both groups, demonstrating potential our suggested framework as useful diagnostic tool various clinical contexts. work clarifies effectiveness techniques, concentrating on crucial job binary classification identification. results open up new avenues investigation field medical diagnosis addition providing insights real-world applications sophisticated machine learning. The study highlights ensemble approach's encouraging benefits, showing it may strengthen precision advance decision-making.

Language: Английский

---

Face Mask Detection and Alert System

Published: May 2, 2024

Language: Английский

---

Advanced frequent itemset mining algorithm (AFIM)

Elsevier eBooks, Journal Year: 2024, Volume and Issue: unknown, P. 187 - 201

Published: Oct. 1, 2024

Language: Английский

---

K-means clustering over distributed environment: A review

Elsevier eBooks, Journal Year: 2024, Volume and Issue: unknown, P. 173 - 185

Published: Oct. 1, 2024

Language: Английский

---