This article curates over 20 machine‑learning cheat sheets—including algorithm overviews, Python libraries, and essential mathematics—compiled from various online sources and illustrated with downloadable diagrams, offering a comprehensive reference as of June 1 2017.
This article describes how Meituan‑Dianping leverages deep learning, especially the Wide & Deep model, to improve its recommendation system by addressing business diversity, user context, feature engineering challenges, optimizer and loss function choices, and presents offline and online experimental results showing significant CTR gains.
This article surveys the rapidly evolving JavaScript machine‑learning ecosystem, introducing ten libraries—from Brain and Synaptic to DeepForge—detailing their capabilities, usage contexts, and example demos, and showing how developers can experiment with AI directly in web browsers.
This article reviews a KDD 2016 study that combines the Regularized Greedy Forest algorithm with transfer learning to accurately predict hard‑disk failures in data centers, addressing challenges like irrelevant SMART attributes, imbalanced data, and model portability across disk models.
This guide outlines a step‑by‑step learning roadmap for ordinary programmers aiming to become data scientists, covering essential math, statistics, machine learning fundamentals, feature engineering, deep learning resources, open‑source tools, and practical project advice to navigate the AI field effectively.
This article reviews YouTube’s recommendation system research from 2008 to 2016, detailing four development stages—user‑video graph walks, video‑video graph walks, search‑based methods with collaborative filtering, and deep neural networks—highlighting key algorithms, system architectures, and experimental results.
The article introduces Weiflow, a dual‑layer DAG‑based machine‑learning workflow framework designed for Sina Weibo, and explains how its modular XML configuration, Scala implementation, and integration with Spark, TensorFlow, Hive, Storm, and Flink improve development efficiency, scalability, and execution performance across the entire ML pipeline.
Alibaba’s 2016 Double‑11 e‑commerce search team unveiled the CLOES cascade ranking algorithm, which reduces CPU usage by about 45%, lowers average search latency from 33 ms to 24 ms, and boosts GMV by nearly 1%, with detailed offline and online validation presented in a KDD‑2017 paper.
Alibaba's senior algorithm expert Yang Hongxia explains how the company fuses massive, heterogeneous data sources into a unified platform, builds automated tag‑production pipelines and large‑scale distributed knowledge graphs, and applies these technologies to drive smarter business decisions and AI‑enabled services.
The article presents a comprehensive overview of DianRong Net's in‑house machine‑learning platform built on Spark, its workflow, pain points it addresses, risk‑control case studies using graph mining, and practical tips for improving model performance through data, algorithms, hyper‑parameter tuning and ensemble methods.
The article explains Gartner’s definitions of Business Intelligence and Artificial Intelligence, compares their core capabilities, discusses whether AI is over‑hyped, and illustrates how AI‑driven solutions differ from traditional BI across marketing, public safety, finance, and education sectors.
Tencent's social advertising team hosts an algorithm contest for college students, leveraging big data and machine learning to develop innovative solutions for social advertising scenarios, inviting participants to submit algorithmic approaches to real-world advertising challenges.
The article explores how machine learning models emulate human learning by converting diverse real‑world descriptions into numerical features, illustrating concepts such as one‑hot encoding, supervised, unsupervised, and reinforcement learning, and emphasizing the importance of mapping inputs to outputs for intelligent systems.
This article explains how applying machine‑learning techniques such as EWMA and periodic‑based amplitude analysis creates dynamic thresholds that dramatically reduce false alerts in monitoring systems while improving detection of genuine anomalies.
This guide outlines a smooth, step‑by‑step learning path for busy programmers with a bachelor's degree, covering AI fundamentals, essential mathematics, recommended courses, practical projects, deep‑learning resources, open‑source tools, and strategies to stay motivated and succeed in the field.
The paper “Optimized Cost per Click in Taobao Display Advertising” introduces a novel two‑level OCPC smart bidding algorithm that jointly optimizes advertiser ROI, user experience, and platform revenue, presents detailed mathematical formulations, offline and online experiments showing significant gains in GMV, CTR, CVR, and RPM across single‑item and banner ad placements.
In this interview, a solo female participant from Shanghai Jiao Tong University shares her experience, challenges, and technical insights—including feature engineering, memory management, and model tuning with XGBoost and LightGBM—gained while competing in the Tencent Social Ads algorithm contest.
The article explains how web crawlers (spiders) collect data for knowledge graphs, covering core tasks, major challenges, crawler features, new‑link expansion, storage design, link‑selection scheduling strategies, and the role of large‑scale data mining and machine learning in optimizing crawl efficiency.
This article examines Toutiao’s rapid rise as a personalized news platform, detailing its AI‑driven recommendation pipeline, web‑crawling infrastructure, similarity‑matrix algorithms, A/B testing, and the role of human moderation in delivering highly targeted content to billions of users.
This article surveys eleven forward‑looking predictions—from cloud computing eclipsing Moore's law to IoT security challenges, video‑centric web experiences, AI‑driven features, evolving UI design, autonomous transport, legal constraints, and the rise of containers—highlighting how programmers must adapt to stay competitive.
Meituan‑Dianping’s travel search team tackles cross‑region queries and noisy data by iteratively refining a four‑step, case‑driven pipeline that classifies intent, segments queries, ranks results with distance and term‑importance models, and employs multi‑stage, parallel recall to steadily boost purchase rate, CTR, and user satisfaction.
This article details the design and implementation of a live‑stream video quality monitoring platform that combines subjective labeling, machine‑learning models, and a multi‑layer architecture to automatically evaluate and alert on video quality across thousands of streams.
This article explains the technical background of click‑through‑rate (CTR) prediction, critiques traditional linear models, introduces Alibaba’s Mixed Logistic Regression (MLR) algorithm with its advanced features and large‑scale distributed implementation, and reviews its successful deployment and remaining challenges in advertising systems.
Suning’s smart replenishment system showcased at CES Asia 2017 leverages big‑data analytics and machine‑learning models—linear regression, random forest, and XGBoost—to predict sales, optimize inventory across multiple warehouses, and shift retail from traditional B2C to a data‑driven C2B approach.
In this article, the author shares their approach to data analysis for a Tencent competition, including daily data partitioning, feature engineering techniques, business understanding, and model selection strategies.
JD.com's JData Algorithm Competition concluded with record-breaking participation, awarding 600,000 yuan in prizes and offering special employment opportunities to top performers, while showcasing how AI and big data enhance e-commerce experiences.
The weekly roundup covers AlphaGo’s retirement and release of self‑play games, Tencent’s surge as Asia’s top startup investor, Microsoft’s migration of Windows code to Git, WeChat mini‑program enhancements, Alibaba’s retail partnership, AI insights from Siri’s co‑founder and Facebook’s new NMT model, Xamarin Live Player’s impact, and Git 2.13 security improvements.
The article describes how a three‑stage algorithm platform was designed and implemented to automate model deployment, unify feature processing, and provide service‑oriented model evaluation, debugging, and monitoring for machine‑learning applications in a large e‑commerce environment.
The article reviews the surge of AI interest sparked by AlphaGo and Master, explains why strong mathematics—especially linear algebra, probability, statistics, calculus, and optimization—is crucial for AI practitioners, and provides curated free online courses, textbooks, and resources to help beginners master these subjects.
The JD Data Algorithm Competition aims to help shoppers overcome decision paralysis by using advanced algorithms to match users with products they'll love, featuring diverse participants from students to professionals competing for prizes and potential real-world implementation.
This article provides a comprehensive introduction to Kaggle, covering its history, competition formats, participation rules, public and private leaderboard mechanics, and a step‑by‑step workflow that includes data analysis, cleaning, feature engineering, model training, validation, hyper‑parameter tuning, ensemble techniques, and automation frameworks for successful data‑mining contests.
Meituan’s demand‑side platform combines real‑time bidding with a two‑service architecture—RecServer for multi‑scenario ad recall and PredictorServer for CTR and conversion‑value prediction—leveraging behavior, location, collaborative‑filtering and matrix‑factorization features, logistic‑regression and GBDT models, and continuous A/B and metric monitoring to optimize ROI.
This article introduces regression concepts, explains linear, locally weighted, and ridge regression methods, demonstrates their mathematical foundations, provides Python implementations, and discusses model evaluation techniques to help readers choose the appropriate regression approach for their data.
The article presents a comprehensive 2017 overview of how Quora applies machine learning across question understanding, answer ranking, personalization, content quality, and ad optimization, detailing the models, libraries, and platform infrastructure that power these systems.
Alibaba unveiled its AI‑driven design platform Luban, which leverages massive data and machine‑learning algorithms to automatically generate high‑quality, personalized graphics at double‑digit scale, reshaping design workflows, boosting efficiency, and prompting a strategic shift for designers in the new commercial era.
This article breaks down the four essential elements of a model project—business, features, data, and model—explains their priority order, and provides practical guidance on aligning technical choices with core business goals to achieve efficient, high‑impact outcomes.
Facebook evaluates its Newsfeed recommendation quality through three pillars—machine-learning model metrics like AUC, extensive product data KPIs such as DAU and interaction rates, and user-survey feedback—while maintaining long-term backtests and emphasizing the risks of relying on a single metric.
This article defines artificial intelligence as a branch of computer science that aims to simulate and extend human intelligence, outlines its key research areas such as robotics and natural language processing, and presents a detailed visual timeline from TalkingData that charts the technology’s major milestones.
This tutorial introduces TensorFlow fundamentals—including computational graphs, sessions, tensors, variables, placeholders, and feed_dict—explains why TensorFlow is popular, and walks through a complete MNIST handwritten digit recognition example with code snippets and visualizations.
Neural Complete, an open‑source project by Pascal van Kooten, trains a neural network to auto‑complete the code of another neural network using LSTM and Keras, demonstrating AI‑driven metaprogramming that could accelerate software development, research, and numerous future applications.
This tutorial walks through using scikit‑learn's LinearRegression on the classic diabetes dataset, covering data description, model training with fit(), making predictions, evaluating performance, and code optimizations, all illustrated with clear output images and plots.
The talk demonstrates how Meipai, a leading Chinese short‑video community, leverages large‑scale text mining and machine‑learning techniques—ranging from anti‑spam filtering to AI‑enhanced search—to enrich captions, comments, and metadata, improve user experience, and inspire further research on text data in video platforms.
The fourth Meitu Internet Technology Salon showcased practical AI and machine learning uses, highlighting Meipai’s text‑anti‑spam, hot‑topic detection, sentiment analysis and personalized video search, while Baidu demonstrated ML‑driven business intelligence tools for multi‑source data mining, user profiling, and intelligent enterprise and HR management.
This article recounts Wu Guanlin’s presentation on the evolution of Vipshop’s personalized recommendation system, detailing the technical challenges of real‑time predictions, the three generations of architecture, the four‑stage recommendation engine, and the VRE platform’s design for scalability and low latency.
This tutorial revisits a K‑Means clustering case on basketball player statistics, walks through the Python implementation using scikit‑learn, demonstrates Matplotlib visualizations, offers plotting optimizations, and addresses common Spyder IDE issues for data‑science workflows.
This tutorial walks you through installing Anaconda, introduces essential Python data‑science packages, explains clustering versus classification, details the K‑means algorithm, and demonstrates a complete K‑means analysis on a basketball player dataset using Python code and visualizations.
The article reviews the major software development trends of 2016—including progressive web apps, chatbots, evolving front‑end frameworks, cloud adoption, and machine‑learning breakthroughs—while highlighting key programming languages, databases, tools, and technologies to watch in 2017.
The article explores how recommendation technologies powered by machine learning permeate everyday platforms—from e‑commerce and video services to social media and news apps—detailing the data they collect, the algorithms they employ, and the limits of personalization in unpredictable human scenarios.
The article outlines Meituan‑Dianping’s tourism recommendation system, detailing its evolution from simple hot‑sale recall to sophisticated decay‑based, GPS‑aware, collaborative filtering and XGBoost reranking pipelines, the four‑layer architecture supporting dozens of travel scenarios, and future plans to broaden recall, adopt deep models, and expand multimodal travel recommendations.
This interview with Alibaba researcher Xu Yinghui reveals how the company built large‑scale deep reinforcement learning systems for search, recommendation, logistics and online advertising, detailing team structures, technical breakthroughs, training challenges, and future directions such as multi‑agent learning and GAN integration.
The article outlines the key competencies, responsibilities, and career development advice for data professionals across the entire data stack—from building big‑data platforms and data warehouses to visualization, analysis, algorithm engineering, and deep‑learning applications—emphasizing the importance of creating business value with data.
This article introduces five highly rated Python libraries—PyWren, Tfdeploy, Luigi, Kubelib, and PyTorch—that streamline data handling, cloud execution, workflow orchestration, and GPU acceleration, helping machine‑learning engineers boost productivity and tackle complex projects more efficiently.
This article explains the concept and construction of user personas, explores the essence and algorithms of recommendation systems, compares movie and e‑commerce scenarios, and details NetEase Yanxuan's practical CTR‑based recommendation model with extensive feature engineering.
The discussion outlines practical anti‑spam strategies—including text length limits, keyword replacement, trie‑based data structures, AC automata, Bayesian and vector‑similarity algorithms, and PHP extensions such as libdatrie—while also sharing performance metrics and resource links for implementing robust content filtering systems.
This article provides a concise overview of sixteen popular Python machine‑learning libraries—including scikit‑learn, NLTK, Theano, and Orange—detailing their main features, typical use cases, and where to find their project pages, making it a handy reference for data‑science practitioners.
This tutorial walks through setting up a Python 2.7 machine learning environment with scikit-learn, installing required libraries, loading web traffic data, cleaning NaN entries, visualizing the data, performing a linear regression using SciPy's polyfit, and evaluating the model's fit.
A diverse group of practitioners share recommendations for books, technologies, real‑world use cases, and practical challenges when learning and applying big‑data processing, covering Hadoop, Spark, data visualization, ETL, and the relationship between data, algorithms, and business value.
This guide introduces beginners to core machine learning concepts, covering feature design, supervised and unsupervised methods such as perceptron, logistic regression, decision trees, LDA, and ensemble techniques like bagging and boosting, while explaining model evaluation, overfitting, and practical optimization strategies.
Alibaba leveraged reinforcement learning, highlighted by MIT Technology Review’s 2017 breakthrough list, to transform its e‑commerce search and recommendation systems during Double 11, deploying large‑scale online and batch training pipelines, dynamic market segmentation, and real‑time decision models that boosted click‑through rates by up to 20 %.
Alibaba’s merchant division introduced a scalable graph‑embedding approach for its “thousands‑of‑people‑one‑face” recommendation module, enabling personalized product suggestions within sparse shop data, improving click‑through rates by 30% and conversions by 60%, and presenting theoretical insights validated at AAAI 2017.
Meituan Delivery’s rapid expansion across multiple categories relies on detailed user profiling and machine‑learning models—such as high‑potential customer prediction, churn risk regression and Cox survival analysis—to personalize acquisition, retention, and scenario‑based cross‑selling, while addressing sparse behavior, unstructured data, and geographic context challenges.
This article shares practical engineering experiences for deploying machine learning models in various scenarios—real‑time low‑volume predictions via Rserve or Python‑httpserve, high‑throughput real‑time serving with PMML‑wrapped Java classes, and offline batch predictions using simple shell scripts—detailing tools, performance considerations, and implementation steps.
This article reviews the fundamentals of cross-domain recommendation, explains the limitations of single‑domain personalized recommendation, surveys existing collaborative‑filtering, transfer‑learning, and knowledge‑based methods, and introduces two novel tensor‑factorization and bilinear multilevel models that achieve superior performance on real datasets.
An illustrated guide explores the recommendation algorithms powering e‑commerce platforms, drawing on Xavier Amatriain’s CMU Machine Learning summer school lectures to explain collaborative filtering, content‑based, and hybrid approaches, their practical implementations, and the impact on user experience and sales.
This article explains how Tencent built a billion‑scale, generalized recommendation system by designing a reusable algorithm library, deploying a low‑latency, highly available real‑time streaming platform (R2), and offering a cloud‑based recommendation engine that simplifies integration for internet businesses.
This article shares practical engineering experiences for deploying machine learning models in production, covering three typical scenarios—real‑time small data, real‑time large data, and offline predictions—and detailing how to use PMML, Rserve, Spark, shell scripts, and related tools to meet performance and operational requirements.
The article offers a comprehensive overview of the 2016 NIPS conference in Barcelona, detailing its history, attendance, Hulu’s contributions as presenters and reviewers, key tutorials, invited talks, award-winning papers, symposium highlights, and the broader impact of deep learning and AI advancements.
The article explains voiceprint (speaker) recognition technology, covering its biological basis, 1:1 verification versus 1:N identification, content‑related versus content‑independent approaches, key acoustic features such as MFCC, the iVector framework, system workflow diagrams, and its use in an Alibaba security challenge.
The article explains that while machine learning has advanced with abundant courses, textbooks, and frameworks, engineers still face hard debugging problems due to algorithmic, implementation, data, and model dimensions, leading to exponential difficulty and long feedback loops that demand intuition and systematic testing.
The paper presents a generic feature‑production framework for Meituan’s food‑delivery ranking system that abstracts statistical feature generation, storage, retrieval, and online loading into configurable dimensions, metrics and operators, enabling developers to add new features with minimal code and dramatically speeding up machine‑learning model iteration.
In this Alibaba Tech Forum talk, senior expert Wei Hu explains how machine learning and big‑data technologies are used to empower merchants with personalized storefronts, intelligent posters, and AI‑driven headlines, boosting their efficiency and sales performance.
At Alibaba's 2016 Double 11 Tech Forum, researcher Xu Yinghui presented how online AI technologies enhance search and recommendation on the e‑commerce platform, turning massive user behavior data into actionable insights that improve traffic allocation and maximize welfare for consumers, sellers, and the platform.
This article presents a comprehensive overview of Ctrip's Aegis risk control system, detailing its modular architecture, rule engine, data service layer, Chloro analytics platform, and future directions, while highlighting the use of streaming, big‑data processing, and machine‑learning models for real‑time fraud detection.
This article explains how to apply machine learning to real‑time distributed processing, covering model basics, data collection, parameter learning, topology design, message flow control, and the Hurricane framework architecture.
This article provides a comprehensive introduction to data mining, covering its definition, goal setting, data sampling, exploration, preprocessing, pattern discovery, model building, evaluation methods, and the main analytical techniques such as classification, regression, clustering, association rules, feature and deviation analysis, and web mining.
This article outlines a four‑stage learning pathway for programmers—from initial exposure to advanced mastery—detailing the goals, recommended resources, and practical activities for each phase, helping readers identify their current level and plan concrete steps toward becoming proficient in machine learning.
This article reviews a comprehensive big‑data lecture covering traditional databases, Hadoop ecosystems, commercial big‑data platforms, computing models, analysis techniques, visualization, and leading vendors, highlighting how these technologies shape today’s data‑driven enterprises.
This article outlines a service‑oriented architecture for a high‑availability personalized recommendation platform, detailing the front‑end, back‑end, crawler, user‑profile modeling, data collection from logs and client events, and processing pipelines using technologies such as Node.js, Python, RabbitMQ/Kafka, MongoDB and TensorFlow.
This article explains how recommendation systems can use item‑level features instead of user ratings, illustrating the approach with Pandora's music‑gene project, detailing feature selection, scoring, distance calculations, standardization, and classification techniques across music, athlete, Iris, and automobile datasets.
This article walks through the fundamentals of recommendation systems, explaining collaborative filtering and various similarity measures—including Manhattan, Euclidean, Minkowski, Pearson correlation, and cosine similarity—while discussing their suitability for dense, sparse, or biased rating data and introducing K‑Nearest Neighbors for practical implementation.
This article describes Ctrip's challenges with web security, evaluates hardware and commercial cloud WAF shortcomings, and presents a low‑cost, low‑risk cloud‑based WAF solution that leverages DNS redirection, closed‑loop rule management, Lua/Tengine deployment, supervised machine‑learning log analysis, and big‑data streaming for real‑time threat detection and mitigation.
This article introduces a programmer‑focused guide to data mining, explaining how personalized recommendation systems work, illustrating historical and modern examples, and outlining the book’s practical Python‑based approach to mastering big‑scale data analysis.
An introductory guide for machine‑learning beginners that covers essential algorithms—including perceptrons, logistic regression, decision trees, LDA, and ensemble techniques like bagging and boosting—explains feature design, model training, evaluation, and practical tips for avoiding under‑ and over‑fitting.
This article investigates whether the 13‑digit user IDs used in Chinese car‑license lotteries are truly random, revealing how the ID generation, seed‑based selection, and hidden patterns—especially the influential seventh digit—affect outcomes, and demonstrates that simple linear models can achieve an AUC of around 0.8 in predicting winners, while also discussing the system’s opacity across major cities.
The article reviews the evolution and challenges of deep learning, outlines major commercial platforms such as DistBelief, COTS, and Adam, compares open‑source frameworks like MXNet, TensorFlow and Petuum, and highlights their architectures, performance metrics, and diverse applications ranging from image recognition to recommendation systems.
This article describes Qunar's personalized demand prediction system for the "Guess You Like" card, detailing how user‑demand associations are mined via rule engines, collaborative filtering, LBS and manual rules, and how ranking models evolve from subjective Bayes to RankBoost and LambdaMart, with experimental evaluation and future LSTM plans.
This article explores how historical monitoring alerts can be analyzed and predicted using linear regression, guiding operations engineers to preprocess data, build regression models, and forecast future alert trends to reduce manual alarm handling and improve system stability.
This article explains how Alibaba combines big‑data analytics, machine learning, and mechanism‑design game theory to create a recommendation system that removes incentives for merchants to generate fake orders, improving fairness and user experience on e‑commerce platforms.
This article provides a light‑hearted yet thorough overview of deep learning, covering neural network fundamentals, layer construction, back‑propagation, ResNet shortcuts, encoder‑decoder structures, PU‑learning for unlabeled data, GPU acceleration, and practical advice on data size, frameworks, and deployment in financial scenarios.
This article presents a series of slides that explain the recommendation algorithms commonly used in e‑commerce, drawing on concepts taught by Xavier Amatriain at Carnegie Mellon University's Machine Learning summer school.
This article presents a comprehensive overview of hotel search ranking, covering problem definition, the distinction between ranking and probability estimation, handling position bias, detailed feature engineering, the AnyBoost linear boosting model, offline evaluation methods, and observed online performance improvements.
This article describes how Qunar replaced a low‑click hot‑words card with a personalized “You May Like” recommendation card, detailing data collection, rule‑based and collaborative‑filtering association methods, learning‑to‑rank models (subjective Bayes, RankBoost, LambdaMart), training‑sample strategies, online experiments, evaluation metrics, and future plans including LSTM‑based sequence modeling.
This article provides a humorous yet informative, non‑mathematical guide to deep learning, covering neural network basics, layer addition, training methods, back‑propagation, unsupervised pre‑training, regularization, ResNet shortcuts, GPU computation, framework choices, and practical advice for applying deep learning to industrial data.
At Qunar, the data center algorithm team developed an automatic ticket classification system that combines Support Vector Machine with word2vec embeddings to handle high‑dimensional, low‑sample text data, achieving 89% accuracy and 80% recall while outlining the full machine‑learning pipeline from feature extraction to deployment.
This article explains what data is, classifies it into structured, semi‑structured and unstructured forms, describes data mining, databases, data warehouses, the full data lifecycle, and surveys the big‑data ecosystem including storage, batch and real‑time processing, resource scheduling, and visualization technologies.
The article explores the author's experiments combining artificial intelligence with typography, detailing the development of algorithms that measure font attributes, compute similarity scores, and generate rule‑based design systems, while reflecting on the challenges, inspirations, and future possibilities of AI‑driven font selection and design.
This article describes how to design and implement a closed‑loop data platform using Python, Java, and Spark stacks, covering data acquisition, structuring, mining, visualization, real‑time processing, and deployment with Docker, ELK, Kafka, and cloud services, illustrated by three industry case studies.
The article describes how Nuomi analyzes refund and complaint data with machine‑learning and big‑data techniques, extracts features for single‑ and multi‑store scenarios, builds decision‑tree models with regional adjustments, and creates an online workflow to promptly intervene on merchants that fail to serve customers.
Xiaomi senior R&D engineer Deng Wenjun shares the evolution of Xiaomi's internet‑finance risk‑control system, describing early rule‑based limits, the adoption of Drools for fast rule deployment, data‑driven modeling with random‑forest classifiers, and ongoing challenges in scalability, latency, and privacy.
The article details Xiaomi's evolution of internet‑finance risk control—from early limit and frequency rules that cut bad‑debt by a third, through adopting the Drools rules engine for rapid deployment and gray‑release, to leveraging random‑forest machine‑learning models and extensive user profiling that reduced fraud by roughly 40%, while addressing privacy and operational challenges.
