International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 11 Issue: 05 | May 2024
www.irjet.net
p-ISSN: 2395-0072
Recipe Renaissance: Leveraging Deep Reinforcement Learning for Food Recommendations Systems in Culinary Innovation. Neha Tyagi1, Vijay Maheshwari2 1M.tech [Department of Computer Science &Engineering] Shobhit University, Meerut .Uttar Pradesh. 2Associate Professor [Department of Computer Science &Engineering] Shobhit University, Meerut .Uttar Pradesh.
---------------------------------------------------------------------***--------------------------------------------------------------------creating the most effective customized meal plans is Abstract - Personalized meal planning systems are
important as well as challenging. Despite their emphasis on meeting nutritional and health needs, traditional meal planning systems have never been able to accommodate people's diverse tastes, lifestyles, or covert eating patterns [1][2][3]. As there is a disconnect in between what we expect and what the system performs often results in partial or no user requirements adherence, thereby affecting the efficacy of the solutions systems are producing [1][4]. Therefore, personalized meal planning systems are undergoing evolution to better accommodate these diverse needs day and night. The objective for our research project is provided by this introduction, which tells how we can predict the interdependent variables between user satisfaction, following suggested meal plans, and the incorporation of user preferences and nutritional data into these creative systems. This study reveals the details of customized food recommendations and the resulting benefits for encouraging healthier eating habits and improving general well-being by diving into the complicated phenomena of nutritional value and taste.
important in addressing individual nutritional needs, health concerns, and dietary choices. However, challenges exist when effectively accommodating diverse user requirements, resulting in low user preference adherence and satisfaction. This research suggests a novel approach that combines the power of collaborative filtering (CF) and deep reinforcement learning (DRL) by applying the Deep Deterministic Policy Gradient (DDPG) algorithm to effectively handle these obstacles completely. Nutrition, health, preferences, culture, flavor, choices, and dietary restrictions factors like these usually influence eating habits. Traditional systems find difficulties in matching meal suggestions with the user's lifestyle, preferences, and individual dietary requirements or restrictions, owning to less-than-ideal user interactions. Our framework leverages CF insights into user preferences, and dietary restrictions, and then applies DRL's adaptive decisionmaking based on users' interactions as feedback. Using the DDPG algorithm to integrate CF with DRL, our approach effectively caters to most of the user requirements and preferences. DRL principles like Markov decision processes (MDPs) are essential for interactive meal recommendations. A reward-shaping feature is included offering multi-criteria decision-making in producing personalized meal plans for the user. Through this research study, we've compared the effectiveness of our proposed integrated methodology against conventional methods. This study sheds light on the possibility of CF-DRL integration in providing individualized meal ideas, promising to improve user experiences and health goals in the best possible way.
Customized nutrition that considers an individual's needs, interests, and health circumstances is what personalized meal planning is all about—it's not simply a cuisine craze. Still, there are obstacles to overcome to achieve this degree of customization. When it comes to delivering consumers with recommendations that are too generic or generic, traditional meal-planning systems frequently fall short. The result is that it is difficult for consumers to adhere to these customized strategies, which lowers engagement rates. Though customized nutrition makes a lot of promises, it is frequently difficult to get the intended results.
Key Words: personalized meal planning, collaborative filtering, deep reinforcement learning, DDPG algorithm, user preferences, dietary choices, nutritional needs, health considerations.
To address these obstacles, this research presents an innovative solution that combines collaborative filtering (CF) with deep reinforcement learning (DRL) through the utilization of the Deep Deterministic Policy Gradient (DDPG) algorithm. The combination results in a significant advancement in personalized meal preparation, providing a comprehensive structure that seamlessly integrates datadriven analysis with flexible decision-making processes.
1. INTRODUCTION In the present digital world, we are living in, the intermixing of culinary exploration and technological advancements has resulted in numerous food recipe platforms, therefore people engage with them differently. Central to this culinary evolution is the creation and utilization of integrated food recommendation systems, seamlessly mixing the domains of flavor and health to provide customized culinary journeys that cater to personal tastes and dietary requirements. The analysis of all these interrelated elements shows that
© 2024, IRJET
|
Impact Factor value: 8.226
Our proposed methodology gains useful insights into users' cultural practices, preferences, and dietary restrictions using the basic approach which is collaborative Filtering. These insights serve as the foundation upon which our Deep
|
ISO 9001:2008 Certified Journal
|
Page 1777