# Statistics Statistics is a mathematical discipline concerned with collecting, organizing, analyzing, interpreting, and presenting quantitative data. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Naoki Egami's Research on Political Methodology and External Validity DevFeed: [Naoki Egami's Research on Political Methodology and External Validity]() Original publisher: [Read original article]() Author: Peter Dizikes | MIT News Published: 2026-09-16T04:00:00Z Content type: article Language: en Sources: [MIT AI News]() Topics: [Statistics](), [Data Science](), [AI Research]() Tags: [artificial-intelligence](), [external-validity](), [faculty](), [idss](), [mit-political-science](), [mit-schwarzman-college-of-computing](), [naoki-egami](), [political-methodology](), [political-science](), [profile](), [research](), [school-of-humanities-arts-and-social-sciences](), [science](), [social-sciences](), [statistics](), [technology-and-society](), [voting-and-elections]() ### AI overview An MIT profile of political scientist Naoki Egami, whose research examines research methodology, external validity, and the mathematical and statistical challenges of studying civic and political phenomena. It also discusses his work on the use of AI tools in research. ### Source excerpt Naoki Egami has become a standout in political methodology, helping refine tools that give scholars durable results. ## Automating EDA With fg-data-profiling DevFeed: [Automating EDA With fg-data-profiling]() Original publisher: [Read original article]() Author: Real Python Published: 2026-09-15T14:00:00Z Content type: tutorial Language: en Sources: [Real Python]() Topics: [Data analysis](), [pandas](), [Python](), [Data Quality](), [Statistics](), [HTML](), [JSON]() Tags: [data-analysis](), [data-quality](), [pandas](), [python](), [statistics](), [tutorial]() ### AI overview A video course on automating exploratory data analysis with fg-data-profiling. It covers generating interactive reports from DataFrames, exporting them to HTML or JSON, analyzing time series, and comparing datasets. ### Source excerpt Automate exploratory data analysis by transforming DataFrames into interactive reports with one command from fg-data-profiling. ## Coherence, Connections, and... Spacetime Crystals DevFeed: [Coherence, Connections, and... Spacetime Crystals]() Original publisher: [Read original article]() Author: Art Anthony Published: 2026-09-09T07:00:00Z Content type: article Language: en Sources: [Nordic APIs]() Topics: [API](), [observability](), [Statistics]() Tags: [2026](), [ai-agents](), [api](), [api-architecture](), [api-ecosystem](), [api-governance](), [api-management](), [api-platform](), [api-strategy](), [artificial-intelligence](), [blog](), [governance](), [observability](), [statistics](), [uptime]() ### AI overview The article discusses API coherence and how API estates can align with organizational goals and change. It covers business gardening, observability, governance, uptime, and usage statistics in the context of API performance and organizational intent. ### Source excerpt Ahead of his Nordic APIs Summit 2026 talk on API coherence, London Stock Exchange Group's Gareth Faull joins us to talk about the art of aligning APIs with organizational intent. Measuring the performance of an API is a relatively straightforward process: ensure observability is in place, follow governance best practices, measure uptime, track usage statistics, ... ## How Bits Database Optimization proves a query rewrite is faster DevFeed: [How Bits Database Optimization proves a query rewrite is faster]() Original publisher: [Read original article]() Author: Alex Weisberger; Nenad Noveljić; Bowen Chen Published: 2026-08-31T00:00:00Z Content type: article Language: en Sources: [Datadog | The Monitor blog]() Topics: [Optimization](), [database monitoring](), [Database](), [synthetic-data](), [datasets](), [cpu](), [IO](), [Security](), [Statistics]() Tags: [cpu](), [data](), [database](), [database-monitoring](), [infrastructure](), [optimization](), [performance](), [security](), [software](), [sql](), [statistics](), [synthetic](), [synthetic-data]() ### AI overview The article explains how Bits Database Optimization validates that a proposed query rewrite is faster. It describes controlled benchmarking with simulated production-like datasets, accounting for cache state, CPU and I/O contention, execution time, and database work. ### Source excerpt Learn how Bits generates synthetic data, measures simulation fidelity, and uses execution time and database work to determine whether an optimization is truly faster. ## Two ways to measure the cumulative impact of experiments DevFeed: [Two ways to measure the cumulative impact of experiments]() Original publisher: [Read original article]() Author: Lukas Goetz-Weiss; Eddie Cai Published: 2026-08-18T00:00:00Z Content type: opinion Language: en Sources: [Datadog | The Monitor blog]() Topics: [experiments](), [Statistics](), [Data Science]() Tags: [analysis](), [experimentation](), [experiments](), [product](), [statistics](), [testing]() ### AI overview This article explains why summing the observed lift from winning experiments overstates cumulative impact because of the winner's curse. It presents two established approaches: a randomized holdout that measures the combined effect directly, and a statistical correction that aggregates existing experiment estimates. Datadog's Cumulative Impact feature applies the correction without requiring a quarter-long holdout and can analyze an entire experimentation program or a filtered team subset. ### Source excerpt Summing individual wins overstates true impact. See two accurate methods, holdouts and Datadog's Cumulative Impact, and how to choose between them. ## Alexander Rakhlin named director of the MIT Statistics and Data Science Center DevFeed: [Alexander Rakhlin named director of the MIT Statistics and Data Science Center]() Original publisher: [Read original article]() Author: Institute for Data, Systems, and Society Published: 2026-08-03T19:50:00Z Content type: news Language: en Sources: [MIT AI News]() Topics: [Statistics](), [Data Science](), [Machine learning]() Tags: [ai-in-statistics](), [alexander-sasha-rakhlin](), [alumni-ae](), [ankur-moitra](), [artificial-intelligence](), [brain-and-cognitive-sciences](), [computer-science-and-technology](), [data-science](), [electrical-engineering-and-computer-science-eecs](), [faculty](), [fotini-christia](), [idss](), [laboratory-for-information-and-decision-systems-lids](), [leadership](), [lids](), [machine-learning](), [mit-bcs](), [mit-faculty-appointments](), [mit-idss](), [mit-leadership](), [mit-schwarzman-college-of-computing](), [mit-statistics-and-data-science-center-sdsc](), [philippe-rigollet](), [richard-dick-larson](), [school-of-engineering](), [school-of-science](), [statistics]() ### AI overview Alexander Rakhlin has been named the next director of MIT's Statistics and Data Science Center, succeeding Ankur Moitra. Rakhlin has been affiliated with the center since 2016 and has led its interdisciplinary doctoral program, overseeing more than 75 successful PhD defenses. ### Source excerpt An expert in machine learning, statistics, and computation, Rakhlin succeeds Professor Ankur Moitra. ## Lorenz and Little: How Much Does Your Tail Cost? DevFeed: [Lorenz and Little: How Much Does Your Tail Cost?]() Original publisher: [Read original article]() Author: Marc Brooker Published: 2026-07-29T00:00:00Z Content type: article Language: en Sources: [Marc Brooker's Blog](), [Marc Brooker's Blog]() Topics: [Latency](), [Statistics](), [Optimization](), [Concurrency]() Tags: [2026](), [concurrency](), [cost](), [latency](), [optimization](), [statistics]() ### AI overview The article explains how tail-latency percentiles contribute to mean latency, concurrency, and service cost. It introduces the empirical Lorenz Curve and uses Little's law to argue that optimizing the tail can substantially reduce concurrency, capacity demand, and lock contention. ### Source excerpt Lorenz and Little: How Much Does Your Tail Cost? Lorenz and Little sounds like hipster burger bar from 2015. It's time for Marc's Amateur Statistics Corner! Today: why I pay a lot of attention to tail latency when optimizing cost. I've written before on the importance of tail latency for customer experience (e.g. in 2026, 2021, and 2021, and 2017). Today, I want to talk about tail latency from the perspective of cost and capacity. Like many system operators, I think about tail latency using percentiles. Here's a question: how much does each of my latency percentiles contributed to the mean latency? Intuitively, the answer is "quite a lot", but can we quantify that? We can! The thing we're looking for is the empirical Lorenz Curve. It directly calculates the answer to the question: given a latency percentile $P$ (e.g. p99=100ms), how much do requests taking shorter than $P$ contribute to the mean latency? (Let's call it $L(P)$ , so the real answer to our question is $1 - L(P)$). Starting from latency samples, the calculation is pretty simple: L = sum(sorted(x)[:k]) / sum(x) (for a set of n latency samples x, and k=p*n). From a vector of quantiles, things get a little more complicated, because we have to choose how to interpolate between the samples and extrapolate out to the maximum. Here I'm interpolating using a power law, which is a little bit of a sin1, but good enough for our purposes. # Calculate 1 - L(p) for a vector of measured quantiles # q - an array of quantiles (e.g. [1, 10, 200, 10000, 20000]) # p - an array of percentiles they're measured at (e.g. [0, 0.5, 0.9, 0.99, 0.999]) # OneMinusL - One minus the empirical Lorenz curve for each of the percentiles def OneMinusL(q, p): ...Show full implementationHide implementation # Calculate 1 - L(p) for a vector of measured quantiles # q - an array of quantiles (e.g. [1, 10, 200, 10000, 20000]) # p - an array of percentiles they're measured at (e.g. [0, 0.5, 0.9, 0.99, 0.999]) # OneMinusL - One minus the empirica ## Six Out of Ten DevFeed: [Six Out of Ten]() Original publisher: [Read original article]() Author: Jon C. Phillips Published: 2026-07-27T00:58:00Z Content type: article Language: en Sources: [Jon C. Phillips]() Topics: [dashboards](), [Server](), [Benchmark](), [Statistics](), [Artificial Intelligence]() Tags: [articles](), [audience-building](), [benchmarks](), [code](), [dashboards](), [design](), [digital-products](), [enterprise](), [marketing](), [music](), [newsletter](), [photography](), [product-engineering](), [server](), [side-projects](), [six-out-of-ten](), [statistics](), [web-development]() ### AI overview The article explains why email delivery and deliverability are different metrics. It compares benchmark results, showing that messages accepted by receiving servers may still reach spam or disappear, and that outcomes vary by sender type, engagement, and measurement window. ### Source excerpt I've been sending email for over a decade. Across the newsletters I run we push roughly three million emails a quarter, every day, to lists in the tens of thousands. ## Waiting for PostgreSQL 20 - Add backend-level lock statistics DevFeed: [Waiting for PostgreSQL 20 - Add backend-level lock statistics]() Original publisher: [Read original article]() Author: depesz Published: 2026-07-06T17:30:55Z Content type: article Language: en Sources: [select \* from depesz;]() Topics: [PostgreSQL](), [Back end](), [Statistics]() Tags: [backend](), [pg-stat-get-backend-lock](), [pg-stat-lock](), [pg20](), [postgresql](), [statistics](), [uncategorized](), [waiting]() ### AI overview A walkthrough of a PostgreSQL patch that adds per-backend lock statistics, including lock wait counts, wait times, and fast-path exceeded counts. The author demonstrates the feature with concurrent psql sessions and discusses its usefulness for diagnosing lock behavior. ### Source excerpt On 30th of June 2026, Michael Paquier committed patch: Add backend-level lock statistics This commit adds per-backend lock statistics, providing the same information as pg_stat_lock. It is now possible to retrieve those stats (lock wait counts, wait times, and fast-path exceeded count) on a per-backend basis. This data can be retrieved with a ... Continue reading "Waiting for PostgreSQL 20 - Add backend-level lock statistics" ## CVE Mid-Year 2026 Check-In: Volume Vertical, Exploitation Rare DevFeed: [CVE Mid-Year 2026 Check-In: Volume Vertical, Exploitation Rare]() Original publisher: [Read original article]() Author: jgamblin Published: 2026-07-01T15:47:10Z Content type: article Language: en Sources: [Jerry Gamblin]() Topics: [Vulnerabilities](), [Security](), [Statistics](), [cisa]() Tags: [2026](), [comparison](), [cve](), [cves](), [report](), [security](), [statistics](), [uncategorized](), [vulnerabilities]() ### AI overview This mid-year review finds that 35,364 CVEs were published in the first half of 2026, up 49.5% from the same period in 2025, while only 85 had entered CISA's KEV list. The article argues that the main challenge is distinguishing exploitable vulnerabilities from the rapidly growing volume of disclosures. ### Source excerpt We are halfway through 2026, so it is time for the mid-year CVE check-in. The short version: the volume curve has gone vertical while exploitation has not. This review covers everything published in the first half of 2026 (Jan 1 - Jun 30, 2026), the volume, the severity, what is actually being exploited, and who ... Read more ## ESET takes part in Operation Endgame to disrupt Amadey and Stealc DevFeed: [ESET takes part in Operation Endgame to disrupt Amadey and Stealc]() Original publisher: [Read original article]() Author: Jakub Tomanek Tomáš Procházka Published: 2026-06-24T12:35:24Z Content type: article Language: en Sources: [WeLiveSecurity]() Topics: [ESET research](), [Malware](), [Threat Research](), [Cybercrime](), [C2](), [High Profile Threats](), [Encryption](), [data](), [Statistics]() Tags: [analysis](), [data](), [encryption](), [eset-research](), [maas](), [malware](), [network](), [research]() ### AI overview ESET Research describes its contribution to Operation Endgame, a coordinated global effort that disrupted the Amadey botnet and Stealc infostealer. The article covers infrastructure tracking, technical and statistical analysis, malware configuration data, command-and-control servers, encryption keys, campaign identifiers, and affiliate-level activity within the malware-as-a-service ecosystem. ### Source excerpt ESET researchers assisted in the global disruption of the Amadey botnet and Stealc infostealer, providing technical analysis, infrastructure tracking, and affiliate-level insights ## ConvApparel: Measuring and bridging the realism gap in user simulators DevFeed: [ConvApparel: Measuring and bridging the realism gap in user simulators]() Original publisher: [Read original article]() Published: 2026-04-09T11:22:00Z Content type: article Language: en Sources: [The latest research from Google]() Topics: [Conversational AI](), [Generative AI](), [LLMs](), [Artificial Intelligence](), [dataset](), [AI Research](), [Statistics](), [data](), [Google]() Tags: [agents](), [ai](), [ai-agents](), [ai-research](), [conversational-ai](), [data](), [evaluation](), [generative](), [generative-ai](), [llm](), [machine-intelligence](), [natural-language-processing](), [research](), [testing](), [training](), [validation]() ### AI overview Google Research introduces ConvApparel, a human-AI conversation dataset and evaluation framework for measuring the realism gap in LLM-based user simulators. It uses Good and Bad agents and validates results through population-level statistics, human-likeness scoring, and counterfactual validation. ### Source excerpt Generative AI ## Who Contributed to PostgreSQL Development in 2025? DevFeed: [Who Contributed to PostgreSQL Development in 2025?]() Original publisher: [Read original article]() Author: Robert Haas (noreply@blogger.com) Published: 2026-01-19T15:29:00Z Content type: article Language: en Sources: [Robert Haas]() Topics: [PostgreSQL](), [Development](), [Code](), [Statistics]() Tags: [2025](), [blog](), [blog-post](), [code](), [contributed](), [contributions](), [development](), [postgresql](), [statistics]() ### AI overview An annual analysis of PostgreSQL code contributions in 2025 counts 266 principal authors and reports the concentration of new lines of code and commits among contributors. The author notes that the methodology has significant limitations and does not capture all important work. ### Source excerpt Here is another annual blog post breaking down code contributions to PostgreSQL itself (not ecosystem projects) by principal author. I have mentioned every year that this methodology has many limitations and fails to capture a lot of important work, and I reiterate that this year as usual. Nonetheless, many people seem to find these statistics helpful, so here they are. Read more " ## Hard-braking events as indicators of road segment crash risk DevFeed: [Hard-braking events as indicators of road segment crash risk]() Original publisher: [Read original article]() Published: 2026-01-13T22:44:00Z Content type: article Language: en Sources: [The latest research from Google]() Topics: [Google](), [Statistics](), [data](), [Android](), [Network](), [Algorithms, Complexity]() Tags: [algorithms-theory](), [analysis](), [android](), [data](), [evaluation](), [google](), [models](), [network](), [product](), [research](), [statistics]() ### AI overview Google Research examines whether hard-braking events collected through Android Auto can indicate crash risk on road segments. Combining anonymized, aggregated hard-braking data with public crash records from California and Virginia, the study finds a statistically significant positive association and identifies hard-braking events as a denser, potentially leading measure for road safety assessment. ### Source excerpt Algorithms & Theory ## 2025 CVE Data Review DevFeed: [2025 CVE Data Review]() Original publisher: [Read original article]() Author: jgamblin Published: 2026-01-01T18:38:43Z Content type: article Language: en Sources: [Jerry Gamblin]() Topics: [Vulnerabilities](), [data](), [Statistics](), [Security](), [Content Management System](), [Linux Kernel](), [web applications]() Tags: [2025](), [analysis](), [cms](), [cve](), [data](), [linux-kernel](), [patch-tuesday](), [report](), [statistics](), [supply-chain](), [trends](), [uncategorized](), [volume](), [vulnerabilities]() ### AI overview This review analyzes 2025 CVE publication data, reporting 48,185 published CVEs, a 20.6% increase from 2024. It highlights stable median CVSS scores, growth in web application and CMS-related flaws, publication clustering around vendor release cycles, and the Linux Kernel as the product with the most listed vulnerabilities. The article recommends prioritizing vulnerabilities by exploitability and automating remediation where possible. ### Source excerpt 2025 set a new baseline with 48,185 published CVEs. While the sheer volume is climbing, the median CVSS score remained surprisingly stable. We are seeing a distinct shift toward web application flaws (specifically in the CMS ecosystem) and a wider distribution of vendors, proving that vulnerabilities are spreading deeper into the supply chain. This massive growth ... Read more ## How we are building the personal health coach DevFeed: [How we are building the personal health coach]() Original publisher: [Read original article]() Published: 2025-10-27T22:36:00Z Content type: article Language: en Sources: [The latest research from Google]() Topics: [Artificial Intelligence](), [Generative AI](), [Google](), [Time Series](), [data](), [Statistics](), [Android](), [iOS]() Tags: [2025](), [ai](), [android](), [gemini](), [generative-ai](), [google](), [health-bioscience](), [ios](), [preview](), [science](), [statistics](), [time-series]() ### AI overview Google describes a Gemini-powered personal health coach designed to provide personalized, adaptive coaching grounded in science and expert oversight. The optional public preview uses Fitbit data to generate health insights and applies numerical reasoning to physiological time-series data such as sleep and activity. ### Source excerpt Generative AI ## Easier Postgres fine-tuning with online\_advisor DevFeed: [Easier Postgres fine-tuning with online\_advisor]() Original publisher: [Read original article]() Author: Carlota Soto Published: 2025-09-16T18:17:18Z Content type: article Language: en Sources: [Blog -- Neon Docs]() Topics: [Databases](), [SQL](), [Statistics](), [real-time]() Tags: [database](), [monitor](), [performance](), [postgres](), [real-time](), [sql](), [statistics]() ### AI overview The article introduces a Postgres extension called online_advisor that analyzes query workloads in real time and recommends indexes, extended statistics, and prepared statements. It monitors query execution data but leaves review and application of the recommendations to the user. ### Source excerpt You've heard this many times before - in order to keep your Postgres database working smoothly, you need to have proper index planning. Too few indexes, and your query performance suffers. Misestimated row counts can also trick the planner into poor choices, and if you're not usi... ## The Modern Data Toolbox DevFeed: [The Modern Data Toolbox]() Original publisher: [Read original article]() Author: Doximity Published: 2025-08-18T00:36:00Z Content type: article Language: en Sources: [Doximity]() Topics: [data](), [Large Language Model](), [Machine learning](), [data-modeling](), [Statistics](), [Machine Learning, Security Attacks]() Tags: [data](), [data-modeling](), [datasets](), [fraud](), [fraud-detection](), [large-language-models-llms](), [machine-learning](), [statistics]() ### AI overview The article explains how to choose among large language models, machine learning, and statistical methods based on data characteristics, goals, scale, and explainability requirements. It argues that complex data problems often benefit from hybrid systems that combine these approaches, illustrating the idea with a multi-layered fraud detection system for payment processing. ### Source excerpt Matching the Tool to the Task A Quick Recap In a previous article, we focused on the strengths of Large Language Models (LLMs), traditional Machine Learning (ML), and statistical methods and recommended 4 key questions to help you choose the right tool for a data solution. Your Data: Is it structured or unstructured? Bounded or unbounded? Your Goal: Do you need prediction, generation, or inference? Your Data Volume: Are you working with massive datasets or limited samples? Your Need for Transparency: Is deep explainability or strict repeatability a requirement? The key takeaway was that LLMs excel at understanding and generating unstructured, unbounded language; ML models are the gold standard for prediction on structured data; and statistics are invaluable for inference and causality, especially with limited data. However, the most complex and valuable real-world problems rarely fit neatly into one box. What if you need to understand unstructured customer feedback and use it to accurately predict churn? This is where hybrid approaches come in, combining the capabilities of each tool to create a system that is greater than the sum of its parts. Below, we present a few examples showcasing how working with hybrid data approaches helps unlock greater value. Hybrid Data Solutions In our experience, the most effective data solutions often emerge from combining multiple data modeling approaches. Rather than viewing LLMs, ML, and statistics as competitors, we recommend considering them as complementary parts of your broader data toolbox. 1. A Multi-Layered Fraud Detection System built using ML, LLM and Statistics Let's consider a high-stakes and regulated environment of a payments processing system. The primary challenge is to detect and block fraudulent transactions in real-time without incorrectly declining legitimate purchases. In addition, the decision-making process should be transparent and auditable. The analytics workhorse of such a system will be a real-time trans ## Revamping Data Science Interviews DevFeed: [Revamping Data Science Interviews]() Original publisher: [Read original article]() Author: Daeus Jorento Published: 2025-05-28T16:00:00Z Content type: article Language: en Sources: [Square Corner Blog RSS Feed]() Topics: [Data Science](), [Statistics](), [data-engineering](), [Machine learning](), [Generative AI]() Tags: [analytics](), [chatgpt](), [data-engineering](), [data-science](), [false-positives](), [generative-ai](), [hiring](), [interviews](), [machine-learning]() ### AI overview This article explains why data science organizations periodically revamp their interview processes. It covers changing business needs, broader problem banks, candidate experience and reputation, and the impact of generative AI tools such as ChatGPT on syntax-heavy technical questions. ### Source excerpt Interviews are not just about improving hiring outcomes - they are about strengthening the entire DS function ## Postgres 18 Beta 1 Introduces Asynchronous I/O and Upgrade and Observability Changes DevFeed: [Postgres 18 Beta 1 Introduces Asynchronous I/O and Upgrade and Observability Changes]() Original publisher: [Read original article]() Author: Heikki Linnakangas Published: 2025-05-08T21:17:13Z Content type: release Language: en Sources: [Blog -- Neon Docs]() Topics: [releases](), [Release notes](), [io\_uring](), [observability](), [Linux](), [Statistics]() Tags: [asynchronous](), [io-uring](), [linux](), [observability](), [performance](), [postgres](), [release](), [release-notes](), [statistics]() ### AI overview Neon contributors describe the Postgres 18 Beta 1 release, highlighting its new asynchronous I/O subsystem, selectable I/O methods including io_uring, retained planner statistics during pg_upgrade, new upgrade options, and observability changes. ### Source excerpt Postgres 18 Beta 1 just shipped. As with previous major releases, this beta includes previews of all features planned for general availability. Read the release notes for the full list of updates, but we're gonna go through some highlights on this post. New in Postgres 18 Asynchr... ## Anomaly Detection in Time Series Using Statistical Analysis DevFeed: [Anomaly Detection in Time Series Using Statistical Analysis]() Original publisher: [Read original article]() Author: Ivan Shubin Published: 2025-04-15T18:45:36Z Content type: tutorial Language: en Sources: [Booking.com Development - Medium]() Topics: [Time Series](), [Statistics](), [Monitoring](), [data](), [Website]() Tags: [analysis](), [anomaly-detection](), [article](), [behavior](), [data](), [grafana](), [metrics](), [monitoring](), [outlier-detection](), [sre](), [statistics](), [time-series](), [web]() ### AI overview This article explains how to build a statistical anomaly detection system for time series data. It describes why static thresholds and comparisons with the same point one week earlier can miss recurring or gradual problems, and introduces standard deviation as a foundational statistical measure. ### Source excerpt Setting up alerts for metrics isn't always straightforward. In some cases, a simple threshold works just fine -- for example, monitoring disk space on a device. You can just set an alert at 10% remaining, and you're covered. The same goes for tracking available memory on a server. But what if we need to monitor something like user behavior on a website? Imagine running a web store where you sell products. One approach might be to set a minimum threshold for daily sales and check it once a day. But what if something goes wrong, and you need to catch the issue much sooner -- within hours or even minutes? In that case, a static threshold won't cut it because user activity fluctuates throughout the day. This is where anomaly detection comes in. What exactly is anomaly detection? Instead of relying on simple rules, it involves analyzing historical data to spot unusual patterns. There are various ways to implement anomaly detection, including machine learning and statistical analysis. In this article, we'll focus on the statistical approach and walk through how we built our own anomaly detection system for time series data from scratch at Booking. The Naïve Approach One common mistake I've seen across different companies and teams is trying to detect anomalies by simply comparing a business metric to its value exactly one week ago. This week vs previous week At first glance, this approach isn't entirely useless -- you can catch some anomalies, as shown in the image above. But is it a reliable long-term solution? Not really. The big flaw is that today's anomaly becomes next week's baseline. That means if the same issue occurs again at the same time next week, it may go completely unnoticed because we're now comparing against a flawed reference point. Outage in previous week That doesn't look right, our simplistic approach doesn't know that last week's data was compromised. Another limitation of this method is that it only considers a single week at a time. But what if perform ## Overengineering an Obsidian dashboard to get better at Marvel Snap DevFeed: [Overengineering an Obsidian dashboard to get better at Marvel Snap]() Original publisher: [Read original article]() Author: Harsh Shandilya Published: 2025-04-01T05:16:52Z Content type: article Language: en Sources: [Posts on Harsh Shandilya]() Topics: [Obsidian](), [dashboards](), [data](), [Statistics]() Tags: [dashboard](), [data](), [games](), [markdown](), [obsidian](), [plugin](), [statistics](), [visualization](), [yaml]() ### AI overview The author describes building an Obsidian-based system to track and visualize Marvel Snap game statistics. The system uses YAML frontmatter, Markdown notes, templates, and the Obsidian DataView plugin to analyze data such as decks, locations, outcomes, and cubes. ### Source excerpt Using data to answer the ultimate gamer question - why am I so bad? ## 2024 CVE Data Review DevFeed: [2024 CVE Data Review]() Original publisher: [Read original article]() Author: jgamblin Published: 2025-01-05T00:04:57Z Content type: article Language: en Sources: [Jerry Gamblin]() Topics: [Statistics](), [data](), [Vulnerabilities](), [vulnerability](), [Protocol (disambiguation)](), [Cisco](), [iOS](), [GitHub](), [Kernel](), [WordPress](), [WordPress Plugins]() Tags: [cisco](), [cve](), [cve-data](), [github](), [kernel](), [numbers](), [open-source](), [packages](), [statistics](), [uncategorized](), [vulnerabilities](), [vulnerability](), [wordpress](), [wordpress-plugins]() ### AI overview This review examines 2024 CVE statistics, reporting 40,009 published CVEs, a 38.83% increase from 2023. It covers publication patterns, CVSS severity scores, CPE records, and CVE Numbering Authorities. ### Source excerpt 2024 brought unprecedented growth in CVE data, so I figured it would be appropriate to start the new year by exploring these statistics and highlighting some of the more intriguing data points. CVEs By The Numbers We ended 2024 with 40,009 published CVEs, up over 38% from the 28,818 CVEs published in 2023. CVEs By Month Month ... Read more ## Likelihood-ratio inference on differences in quantiles DevFeed: [Likelihood-ratio inference on differences in quantiles]() Original publisher: [Read original article]() Author: Miller, Evan Published: 2024-08-05T10:15:00Z Content type: article Language: en Sources: [Evan Miller]() Topics: [Statistics](), [Inference]() Tags: [confidence-interval](), [experimentation](), [statistics]() ### AI overview This paper presents a two-sample hypothesis test and confidence interval for differences in quantiles using a likelihood-ratio test statistic. A conservative version avoids density estimation, while another version uses a density estimator and produces confidence intervals close to nominal coverage. The method can be computed from four order statistics from each sample. ### Source excerpt Quantiles can represent key operational and business metrics, but the computational challenges associated with inference has hampered their adoption in online experimentation. In a new paper, I present a two-sample difference-in-quantile hypothesis test and confidence interval based on a likelihood-ratio test statistic. It can be computed using only four order statistics from each sample. arXiv link: Likelihood-ratio inference on differences in quantiles [Next page]()