Preprint / Version 1

A Machine Learning Based Fashion Recommendation System

How can the implementation of supervised machine learning enhance the efficiency of discovering apparel that suits individual style preferences?

##article.authors##

  • Shailja Tyagi Shailja Tyagi

DOI:

https://doi.org/10.58445/rars.538

Keywords:

fashion, Machine Learning, apparel

Abstract

In recent years, recommendation algorithms that present content based on user past preferences have gained significant popularity on platforms like TikTok, Instagram, and YouTube. These algorithms effectively connect individuals with products they desire, providing value to both users and content creators. This project proposes to create a similar suggestion algorithm for apparel and fashion. This algorithm will utilize supervised machine learning over a large pre-existing dataset of apparel purchases to provide fashion suggestions based on explicitly stated individual preferences. Such an approach can address the challenges of choice overload faced by individuals seeking to curate their wardrobes. Moreover, it has the potential to facilitate retailers by connecting them with customers actively seeking fashion items that align with their unique tastes. In today's world, fashion recommendation systems have become increasingly vital, enhancing the shopping experience and fashion discovery process for consumers. This research paper explores various techniques for developing a fashion-based recommendation algorithm and emphasizes the significance of recommendation systems. It presents an implemented model and evaluates its performance in suggesting apparel based on individual preferences, demonstrating the effectiveness of the proposed approach in enhancing fashion recommendations.

References

‌“Cosine Similarity.” GeeksforGeeks, 2 Oct. 2020, www.geeksforgeeks.org/cosine-similarity/.

“H&M Personalized Fashion Recommendations.” Kaggle.com, www.kaggle.com/competitions/h-and-m-personalized-fashion-recommendations/data?select=transactions_train.csv.

IBM. “What Is the K-Nearest Neighbors Algorithm? | IBM.” Www.ibm.com, 2023, www.ibm.com/topics/knn.

Isinkaye, F.O., et al. “Recommendation Systems: Principles, Methods and Evaluation.” Egyptian Informatics Journal, vol. 16, no. 3, Nov. 2015, pp. 261–273, www.sciencedirect.com/science/article/pii/S1110866515000341, https://doi.org/10.1016/j.eij.2015.06.005.

Karabiber, Fatih. “Cosine Similarity.” Www.learndatasci.com, www.learndatasci.com/glossary/cosine-similarity/.

Dresoo – the Best Online Outfit Creator. dresoo.com/.

Real Python. “Build a Recommendation Engine with Collaborative Filtering.” Realpython.com, Real Python, 10 July 2019, realpython.com/build-recommendation-engine-collaborative-filtering/.

“Recommendation System Using K-Nearest Neighbors |Use Case in Python.” Analytics Vidhya, 20 Aug. 2020, www.analyticsvidhya.com/blog/2020/08/recommendation-system-k-nearest-neighbors/.

SciKit-Learn. “3.1. Cross-Validation: Evaluating Estimator Performance — Scikit-Learn 0.21.3 Documentation.” Scikit-Learn.org, 2009, scikit-learn.org/stable/modules/cross_validation.html.

Stieg, Cory. “What Is a Recommender System in Machine Learning?” Codecademy Blog, 6 Dec. 2022, www.codecademy.com/resources/blog/what-is-recommender-systems-machine-learning/.

Zach. “Leave-One-out Cross-Validation in Python (with Examples).” Statology, 4 Nov. 2020, www.statology.org/leave-one-out-cross-validation-in-python/.

Downloads

Posted

2023-10-02