3830 research outputs found
Sort by
Weighted Hybrid recommendation system using autoencoders
No full textDeciphering user-item interaction data effectively has become critical to improving user experience and driving business growth in the exponentially expanding digital environment. This research is primarily concerned with exploring the possibility of enhancing recommendation systems' accuracy through the use of a weighted hybrid model, By extracting latent features through autoencoders and seamlessly integrating them with item-based filtering, our system aims to capture and predict user preferences with improved precision. In our experiments, the base autoencoder model exhibited an RMSE score of 0.4980, while the base item-based approach registered an RMSE score of 0.2813. Conclusively, the final weighted hybrid model yielded a score of 0.2666913, underscoring the efficacy of combining different models. The results showcased that the model exhibited an RMSE score of 0.2666913 and the corresponding weights were w1(weight for item-based)=0.1 and w2(weight for autoencoder)=0.9. The report initially reviews the current state of recommendation systems and autoencoders, followed by a comprehensive understanding of the autoencoder mathematical framework
Financial performance analysis using machine learning algorithms: post-IPO of Nykaa
No full textThe research looks at how the financial performance of Nykaa which took the IPO in 2021 through 2023 in three years with machine learning models. This work aims to provide more clarity and prediction of the financial situation by using the regression analysis, time series forecasting, and clustering algorithms represented by Python and allowing this project to uncover patterns hidden within Nykaa's financial data. The literature review goes through current studies on machine learning in financial analysis which also deals with research gaps and adds to its practical concerns. The research proposal, through addressing the deficiency and taking advantage of ML-based predictive analytics, seeks to offer result-oriented insights that will be of importance to investors, general observers, and shareholders operating in the digital commerce arena. The study thus looked into the post-IPO finance and stock analysis of Nykaa that is a well-known e-commerce venture in India applying a multidisciplinary approach including clean enrichment, machine learning, and statistical modelling. Already familiarity with Python programming language and SciPy, NumPy, Pandas, Matplotlib, Seaborn, Statsmodels, TensorFlow, and scikit-learn libraries along with ARIMA and LSTM models, the research explore to a Nykaa's stock price precisely and to gain information about its financial health. It carried out the study and received historical stock price data from Yahoo Finance. The data revealed patterns in the growth of Nykaa’s revenues, profitability measures, and cash flow movements. The study used descriptive statistical approaches and visualization methods to uncover critical information about the Nykaa share price fluctuations, capitalization movement patterns, and trading volume developments. Along with this, the trainings of ARIMA and LSTM are quite good in predicting Nykaa's stock prices in the future, where it proves the real applications of machine learning in financing. Therefore, the effect of the IPO of Nykaa in the context of its financial statements' performance provides a good background of how Nykaa's investors, financial analysts, and stakeholders can understand and react responsibly to the financial market dynamics
Influence of playing video games and gaming habits on levels of resilience, self-efficacy, and coping
No full textThis is a quantitative and correlational study that aimed to investigate the relationship between playing video games on psychological traits resilience, self-efficacy and approach/avoidance coping mechanisms, considering levels of game engagement and the frequency and time spent gaming. A total of 190 video game players respondents were recruited through an online survey. The scale Game Engagement Questionnaire (GEQ) measured the levels of players’ engagement in their games, which was later divided into two groups of low and high engagement. Reliable scales were used to measure psychological resilience (BRS), self-efficacy (GSE-6) and approach/avoidance coping mechanisms (BACQ). Frequency and time measures were based on Lemmens et al. (2015). One hypothesis was supported which showed a significant relationship between low game engagement and high resilience as well as high game engagement and low resilience. It is suggested that motive for playing games may have higher influence on psychological traits
Adoption of DevOps in the Software Development Team: Challenges and Recommendations
No full textThe software industry is advancing. Faster time to market is a competitive advantage, but it is challenging to balance quality and speed. While Agile shortens the development cycle, synchronizing the objectives of the development and operations teams is challenging. DevOps as an Agile extension aims to close this gap by encouraging shared accountability. This thesis surveyed 21 IT specialists and yielded 3 significant challenges that the teams are facing. The absence of support from management and administration. DevOps requires learning an extensive toolkit. Accepting the new working style is challenging. DevOps brings an additional burden on the teams. A lot of tasks are repetitive and timeconsuming. Automation smoke testing can be a good solution to improve accuracy and help reduce efforts. Recommendations are provided to assist in resolving the issues. Support for DevOps should come from all organisational levels. Sufficient training and adaptability are necessary to grasp DevOps ideas. Knowledge sharing within development teams
Unveiling customer sentiments in fast fashion: A comparative analysis of customer reviews, recommendations, and ratings
No full textThe fast fashion industry has evolved significantly, driven by globalization, technology, and changing consumer preferences. Online shopping has become prevalent, enabling fast fashion brands to reach customers globally. Understanding customer sentiments is essential for fast fashion brands to stay competitive and meet evolving customer expectations. This study aims to analyze customer sentiments in the fast fashion industry using a real-world dataset. The research involves feature engineering, sentiment analysis, topic modelling, customer profiling, and dashboard creation. The findings revealed that positive sentiments dominate, with customers expressing satisfaction with products and services. Common topics include fashion focus, customer preferences, quality, and price. The study provides valuable insights for fast fashion brands to tailor their offerings, enhance customer satisfaction, and drive loyalty. Future research could explore building an advanced dynamic dashboard incorporating topic modelling and personalization strategies to further improve customer engagement and brand loyalty in the fast fashion industry
In- Hospital Specialist Referral Ordering System For Nurses
No full textMy project leverages the power of the MERN (MongoDB, Express.js, React, Node.js) stack to create a sophisticated and modernized "In-Hospital Specialist Referral Ordering System" aimed at optimizing the referral process within hospital facilities. The development includes various functionalities, such as viewing referrals, selecting, and viewing/adding specialists, and making referrals. The project aims to foster collaboration between primary care providers(nurses) and specialists while ensuring compliance with healthcare regulations. This project will enhance efficiency and accuracy in specialist referrals, highlighting the importance of effective communication in healthcare and the challenges associated with manual referral processes. Moreover, this project assisted me to understand further the creation of a MERN stack development and the importance of frontend and backend. Hopefully this project be an idea for further development, which will help In-hospital referral system, and bridge the gap between primary care providers and specialists, optimizing patient care pathways
Prevalence of sexual content in PC video games
No full textThis report outlines a research project aimed at quantifying the prevalence and emerging themes of sexual content in video games, specifically those available on the Steam platform. The study involves comprehensive data analysis of a pre-existing data set, preprocessing, and the application of Natural Language Processing (NLP) and machine learning techniques to analyse game descriptors, user reviews, and tags. The project's objectives include understanding implications and providing actionable insights for game developers and writers. The student's learning objectives encompass developing proficiency in data analysis, pre-processing, machine learning and application. This research is expected to offer valuable perspectives on content trends within the PC gaming community, guiding future content creation in the industry
The impact of agile project management methodologies on software development and project success rates
No full textThis research has focused on analysing the impacts of agile method in case of project management for software development and its success rates. Considering the differential impact of agile methods on construction projects due to different levels of implementation, this research has applied a survey method to collect data from 74 respondents from the construction sector by using convenience sampling technique. Frequency analysis and statistical analysis have been done in this research and it is found that agile practices can improve project quality with burndown charts in sprints, increasing flexibility, developing team communication guidelines, and others. The statistical outcomes of the results have also supported these findings and presented that agile project management method has a significant impact on the software project development and it helps in improving success rates
Garbage Classification using transfer learning with CNN and Generative Adversarial Networks (GAN)
No full textPredictive Analytics for City Crime Using Machine Learning
No full textThis study delves into the application of machine learning algorithms to predict urban crime, focusing on the dynamic and complex landscape of Los Angeles. Utilising a comprehensive dataset, the research explores the efficacy of various models like LSTM, GRU, SimpleRNN, Prophet, and XGBoost in predicting crime locations and types. The study aims to transition urban safety strategies from reactive to proactive measures by accurately forecasting crime patterns. The models were evaluated based on RMSE, MAE, and R² scores, with the data split into 80:20 and 70:30 ratios for training and testing. Results indicated that while LSTM, GRU, and XGBoost demonstrated high accuracy in spatial predictions, all models faced challenges in accurately predicting crime types, reflecting the multifaceted nature of criminal behaviour. The study highlights the potential of machine learning in enhancing urban safety but also notes the ethical and practical challenges inherent in predictive policing. It underscores the need for further research, especially in improving crime-type predictions and addressing the ethical implications of such predictive technologies. This research contributes significantly to the field of predictive urban crime analytics, offering insights and pathways for future innovations