{"id":4955,"date":"2026-04-01T15:08:08","date_gmt":"2026-04-01T07:08:08","guid":{"rendered":"https:\/\/trg.unhas.ac.id\/datascienceai\/?p=4955"},"modified":"2026-04-09T20:39:20","modified_gmt":"2026-04-09T12:39:20","slug":"smarter-markets-with-ai-how-lstm-and-gru-redefine-stock-price-prediction","status":"publish","type":"post","link":"https:\/\/trg.unhas.ac.id\/datascienceai\/smarter-markets-with-ai-how-lstm-and-gru-redefine-stock-price-prediction\/","title":{"rendered":"Smarter Markets with AI: How LSTM and GRU Redefine Stock Price Prediction"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\"><em>MAKASSAR, Indonesia \u2013 Apr 1 2026\/DSAI\/<\/em> \u2014 Predicting stock prices has always been a complex and risky task, influenced by countless variables and rapid market changes. However, a recent study demonstrates how artificial intelligence (AI), particularly deep learning models, can significantly improve prediction accuracy by learning patterns hidden within financial time-series data.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The research explores the use of two advanced Recurrent Neural Network (RNN) architectures\u2014<strong>Long Short-Term Memory (LSTM)<\/strong> and <strong>Gated Recurrent Unit (GRU)<\/strong>\u2014to forecast stock prices more effectively. Unlike traditional models, these algorithms are designed to remember long-term dependencies in sequential data, making them especially suitable for analyzing stock market trends .<\/p>\n\n\n\n<figure class=\"wp-block-image aligncenter size-large\"><img fetchpriority=\"high\" decoding=\"async\" width=\"1024\" height=\"610\" src=\"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/stock-price-1024x610.png\" alt=\"\" class=\"wp-image-4956\" srcset=\"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/stock-price-1024x610.png 1024w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/stock-price-300x179.png 300w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/stock-price-768x457.png 768w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/stock-price.png 1322w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">The study uses extensive historical data from Yahoo Finance, covering multiple companies over a 12-year period. According to the <em>data description on page 5<\/em>, each company contributes over 15,000 data points, including open, high, low, close prices, and trading volume . Among these, the <strong>closing price<\/strong> is identified as the most critical feature, as it reflects the final consensus of market activity at the end of each trading day (<em>page 6<\/em>) .<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">What makes this research particularly innovative is its <strong>grouped time-series approach<\/strong>. Instead of analyzing each stock independently, the model first learns patterns from individual stocks and then combines them into a unified representation. This allows the system to capture relationships between different stocks\u2014something that traditional models often overlook (<em>page 7<\/em>) .<\/p>\n\n\n\n<figure class=\"wp-block-image aligncenter size-large\"><img decoding=\"async\" width=\"1024\" height=\"443\" src=\"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-1024x443.png\" alt=\"\" class=\"wp-image-4958\" srcset=\"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-1024x443.png 1024w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-300x130.png 300w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-768x333.png 768w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png 1330w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">The researchers developed <strong>four architectural models<\/strong>, ranging from simple direct models to more advanced designs with dimensional reduction, dropout tuning, and stabilization layers (<em>pages 8\u20139<\/em>) . These variations were tested using three evaluation metrics\u2014MAPE, RMSPE, and RMDPE\u2014to measure accuracy, average performance, and risk levels.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The results are striking. According to the <em>evaluation results on page 17<\/em>, the <strong>GRU-based models consistently outperform LSTM<\/strong>, achieving up to <strong>98.48% accuracy during training<\/strong> and <strong>97.37% during testing<\/strong> . Meanwhile, LSTM models demonstrate more stable and consistent predictions, with lower variance across different scenarios.<\/p>\n\n\n\n<figure class=\"wp-block-image aligncenter size-large\"><img decoding=\"async\" width=\"1024\" height=\"644\" src=\"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-8-1024x644.png\" alt=\"\" class=\"wp-image-4957\" srcset=\"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-8-1024x644.png 1024w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-8-300x189.png 300w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-8-768x483.png 768w, https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-8.png 1327w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">This research contributes strongly to multiple <strong>Sustainable Development Goals (SDGs)<\/strong>. It supports <strong>SDG 9: Industry, Innovation, and Infrastructure<\/strong> by advancing AI-driven financial technologies. It also aligns with <strong>SDG 8: Decent Work and Economic Growth<\/strong>, as improved forecasting tools can help investors and businesses make more informed financial decisions. Additionally, it contributes to <strong>SDG 17: Partnerships for the Goals<\/strong>, reflecting collaborative academic research across institutions.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Ultimately, the study highlights an important insight: accurate stock prediction is not just about choosing the right algorithm, but about designing intelligent architectures that capture real-world complexity. By combining LSTM and GRU with grouped time-series learning, this research brings us one step closer to more reliable and practical AI-driven financial systems.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Article Metadata (Source of This Popular Scientific Writing):<\/strong><br>Title: <em>Implementation of Long Short-Term Memory and Gated Recurrent Units on Grouped Time-Series Data to Predict Stock Prices Accurately<\/em><br>Journal: <em>Journal of Big Data<\/em><br>Publisher: Springer Nature<br>Publication Date: 2022 (Volume 9, Article 89)<br>DOI: <a href=\"https:\/\/doi.org\/10.1186\/s40537-022-00597-0\">https:\/\/doi.org\/10.1186\/s40537-022-00597-0<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Authors &amp; Contact:<\/strong><br>Armin Lawi*, Hendra Mesra, Supri Amir<br>Hasanuddin University, Indonesia &amp; collaborating institutions<br>E-mail: <a href=\"mailto:armin@unhas.ac.id\">armin@unhas.ac.id<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This study shows that the future of financial forecasting lies in intelligent collaboration\u2014not just between investors, but between algorithms themselves.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>MAKASSAR, Indonesia \u2013 Apr 1 2026\/DSAI\/ \u2014 Predicting stock prices has always been a complex and risky task, influenced by countless variables and rapid market changes. However, a recent study demonstrates how artificial intelligence (AI), particularly deep learning models, can significantly improve prediction accuracy by learning patterns hidden within financial time-series data. The research explores [&hellip;]<\/p>\n","protected":false},"author":60,"featured_media":4958,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_uag_custom_page_level_css":"","_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[1],"tags":[46,37,38],"class_list":["post-4955","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized","tag-sdg17","tag-sdg8","tag-sdg9"],"acf":[],"uagb_featured_image_src":{"full":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png",1330,576,false],"thumbnail":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-150x150.png",150,150,true],"medium":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-300x130.png",300,130,true],"medium_large":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-768x333.png",768,333,true],"large":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-1024x443.png",1024,443,true],"1536x1536":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png",1330,576,false],"2048x2048":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png",1330,576,false]},"uagb_author_info":{"display_name":"Admin DSAI","author_link":"https:\/\/trg.unhas.ac.id\/datascienceai\/author\/admindatascienceai\/"},"uagb_comment_info":0,"uagb_excerpt":"MAKASSAR, Indonesia \u2013 Apr 1 2026\/DSAI\/ \u2014 Predicting stock prices has always been a complex and risky task, influenced by countless variables and rapid market changes. However, a recent study demonstrates how artificial intelligence (AI), particularly deep learning models, can significantly improve prediction accuracy by learning patterns hidden within financial time-series data. The research explores&hellip;","rttpg_featured_image_url":{"full":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png",1330,576,false],"landscape":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png",1330,576,false],"portraits":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png",1330,576,false],"thumbnail":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-150x150.png",150,150,true],"medium":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-300x130.png",300,130,true],"large":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9-1024x443.png",1024,443,true],"1536x1536":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png",1330,576,false],"2048x2048":["https:\/\/trg.unhas.ac.id\/datascienceai\/wp-content\/uploads\/sites\/76\/2026\/04\/image-9.png",1330,576,false]},"rttpg_author":{"display_name":"Admin DSAI","author_link":"https:\/\/trg.unhas.ac.id\/datascienceai\/author\/admindatascienceai\/"},"rttpg_comment":0,"rttpg_category":"<a href=\"https:\/\/trg.unhas.ac.id\/datascienceai\/category\/uncategorized\/\" rel=\"category tag\">Uncategorized<\/a>","rttpg_excerpt":"MAKASSAR, Indonesia \u2013 Apr 1 2026\/DSAI\/ \u2014 Predicting stock prices has always been a complex and risky task, influenced by countless variables and rapid market changes. However, a recent study demonstrates how artificial intelligence (AI), particularly deep learning models, can significantly improve prediction accuracy by learning patterns hidden within financial time-series data. The research explores&hellip;","_links":{"self":[{"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/posts\/4955","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/users\/60"}],"replies":[{"embeddable":true,"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/comments?post=4955"}],"version-history":[{"count":2,"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/posts\/4955\/revisions"}],"predecessor-version":[{"id":5150,"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/posts\/4955\/revisions\/5150"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/media\/4958"}],"wp:attachment":[{"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/media?parent=4955"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/categories?post=4955"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/trg.unhas.ac.id\/datascienceai\/wp-json\/wp\/v2\/tags?post=4955"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}