[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-118559-en":3,"doc-seo-118559-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},118559,13056703019662,"Evangeline","https://ap-avatar.wpscdn.com/avatar/be000253a8e92610077?_k=1778726343310543188",8,"Research & Report","Machine Learning Predictors for Min-Entropy Estimation","This study investigates machine learning predictors for min-entropy estimation in Random Number Generators (RNGs) used in cryptographic applications. It finds that predictors leveraging sequence correlations primarily estimate average min-entropy rather than min-entropy, and analyzes how the mapping depends on the number of target bits. Using data from Generalized Binary Autoregressive Models, the work shows that hybrid convolutional-recurrent LSTM and GPT-2 transformer models can outperform traditional NIST SP 800-90B predictors in specific scenarios.","Machine Learning Predictors for Min-Entropy Estimation  \nJavier Blanco-Romero, Vicente Lorenzo, Florina Almenares Mendoza, Daniel D´ıaz-Snchez  \narXiv :2406 . 19983v1 [ cs .LG] 28 Jun 2024  \nAbstract—This study investigates the application of machine learning predictors for min-entropy estimation in Random Number Generators (RNGs), a key component in cryptographic applications where accurate entropy assessment is essential for cybersecurity. Our research indicates that these predictors, and indeed any predictor that leverages sequence correlations, primarily estimate average min-entropy, a metric not extensively studied in this context. We explore the relationship between average min-entropy and the traditional min-entropy, focusing on their dependence on the number of target bits being predicted. Utilizing data from Generalized Binary Autoregressive Models, a subset of Markov processes, we demonstrate that machine learning models (including a hybrid of convolutional and recurrent Long Short-Term Memory layers and the transformerbased GPT-2 model) outperform traditional NIST SP 800-90B predictors in certain scenarios. Our findings underscore the importance of considering the number of target bits in minentropy assessment for RNGs and highlight the potential of machine learning approaches in enhancing entropy estimation techniques for improved cryptographic security.  \nIndex Terms—Min-entropy Estimation, Machine Learning Predictors, Random Number Generators, Autoregressive Processes, Generalized Binary Autoregressive Models  \nI. INTRODUCTION  \nTHE security of cryptographic systems often hinges on the  \ngeneration of random values. Although there is a broad spectrum of algorithms and devices used to generate these random values, they are all generically denoted by Random Number Generators (RNGs) . Given the important role that RNGs play in the context of cybersecurity, it becomes evident that rigorous criteria are necessary for evaluating the reliability and performance of an RNG.  \nMultiple approaches are commonly employed to assess the quality of the output of an RNG (cf. [1], [2], [3], [4], [5], [6], etc.) . In this paper the emphasis will be put on:  \n• Entropy tests, as those found in NIST Special Publication 800-90B [7], which estimate the entropy of a noise source based on appropriate samples (cf. [8], [9], [10], etc.) .  \n• Machine Learning models trained with the output of an RNG aiming to guess the bit or set of bits that follow  \nJavier Blanco-Romero is with the Department of Telematic Engineering, Universidad Carlos III de Madrid, Legans, Madrid, 28911, Spain (e-mail: [frblanco@pa.uc3m.es](frblanco@pa.uc3m.es)).  \nVicente Lorenzo is with the Department of Telematic Engineering, Universidad Carlos III de Madrid, Legans, Madrid, 28911, Spain (e-mail: [vilorenz@pa.uc3m.es](vilorenz@pa.uc3m.es)).  \nFlorina Almenares Mendoza is with the Department of Telematic Engineering, Universidad Carlos III de Madrid, Legans, Madrid, 28911, Spain ([e-mail: florina@it.uc3m.es](e-mail: florina@it.uc3m.es)) .  \nDaniel D´ıaz-Snchez is with the Department of Telematic Engineering, Universidad Carlos III de Madrid, Legans, Madrid, 28911, Spain (e-mail: [dds@it.uc3m.es](dds@it.uc3m.es)) .  \na given sequence, which can give an insight into how predictable the output of the RNG is (cf. [11], [12], [13],[14], [15], [16], etc.) .  \nThe fact that the entropy of a given source and the predictability of its output are correlated was already noticed by Shannon [17] . Nevertheless, the link between these two concepts is far from being completely understood, specially if one takes into account the heterogeneity of entropy definitions that can be found in the literature and how much the predictability of the output of an entropy source relies on the predictor being considered. Building on the evidence provided by [18] that the entropy estimators considered by NIST Special Publication 800-90B [7] tend to underestimate min-entropy, an attempt to reinfor","cbCaigQJX5Bn9jYg","https://ap.wps.com/l/cbCaigQJX5Bn9jYg","pdf",3185286,1,17,"English","en",105,"# Introduction\n## Entropy tests for RNG assessment\n## Predictors based on machine learning\n## Relationship between entropy and predictability\n## Predictors for average min-entropy\n## Impact of target bit count\n# Related structure (planned)\n## Literature review\n## Theoretical framework\n## Experimental outline","[{\"question\":\"What problem does the paper address in RNG security evaluations?\",\"answer\":\"It addresses how to assess RNG output reliability by estimating entropy, focusing on min-entropy as a core metric for cryptographic security.\"},{\"question\":\"What does the paper claim about machine learning predictors' entropy estimates?\",\"answer\":\"It argues that predictors using sequence correlations mainly estimate average min-entropy, with min-entropy and average min-entropy linked through a dependence on the number of predicted target bits.\"},{\"question\":\"Which machine learning models are evaluated, and how do they compare to NIST SP 800-90B?\",\"answer\":\"The study evaluates a hybrid convolutional + recurrent LSTM model and a transformer-based GPT-2 model, showing they can outperform traditional NIST SP 800-90B predictors in certain scenarios.\"}]","Machine Learning Predictors for Min-Entropy Estimation | PDF",1785684155,43,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"machine-learning-predictors-for-min-entropy-estimation","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/machine-learning-predictors-for-min-entropy-estimation/118559/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the paper address in RNG security evaluations?","Question",{"text":75,"@type":76},"It addresses how to assess RNG output reliability by estimating entropy, focusing on min-entropy as a core metric for cryptographic security.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What does the paper claim about machine learning predictors' entropy estimates?",{"text":80,"@type":76},"It argues that predictors using sequence correlations mainly estimate average min-entropy, with min-entropy and average min-entropy linked through a dependence on the number of predicted target bits.",{"name":82,"@type":73,"acceptedAnswer":83},"Which machine learning models are evaluated, and how do they compare to NIST SP 800-90B?",{"text":84,"@type":76},"The study evaluates a hybrid convolutional + recurrent LSTM model and a transformer-based GPT-2 model, showing they can outperform traditional NIST SP 800-90B predictors in certain scenarios.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]