#pyod
PyODで時系列データの異常検知を試す
https://zenn.dev/drbobt1122/articles/blog26101701
PyODで時系列データの異常検知を試す
zenn.dev
October 18, 2026 at 5:36 AM
今日のZennトレンド

ClaudeにiOSシミュレー...
https://zenn.dev/eai/articles/dont-let-agents-see-the-ios-sim-screen
AFM(Apple Found...
https://zenn.dev/gritchbreed/articles/ced633f449178f
PyODで時系列データの異常検...
https://zenn.dev/drbobt1122/articles/blog26101701
October 19, 2026 at 12:59 AM
今日のZennトレンド

PyODで時系列データの異常検知を試す
PyODライブラリを用いて、時系列データの異常検知手法7種類を比較検証した内容です。
点異常と部分系列異常の2つのデータ形式に対し、WindowedBridgeやMatrixProfileなどの手法を適用し、パラメータ変更による結果への影響や手法ごとの特性を確認しています。
単に手法の優劣を決めるのではなく、データや目的に応じて適切な検出器を使い分けることの重要性を伝えています。
PyODで時系列データの異常検知を試す
前回は、Raspberry Pi 4に保存したセンサデータを対象に、PyODのSpectral Residualで異常を検出するMCPサーバを構築しました。この話の最後に、データや目的に合わせて異常検知の手法を切り替える旨について、書きました。そこで今回は、PyODに入っている時系列データ向けの異常検出器について、調査し、その結果の違いを確認しました。試したのは次の7種類です。カテゴリ検出器概要W
zenn.dev
October 18, 2026 at 1:03 PM
PyODで時系列データの異常検知を試す
PyODの時系列異常検出器7種を同じ周期データで比べ、点異常と部分系列異常それぞれで効く設定や得意な手法を調べています。

規則的な成分を平滑化して差し引き、残った外れを見ると、規則の中の例外が際立ちます。予測や残差からのずれで異常を見る方法全般に通じる考え方です。
PyODで時系列データの異常検知を試す
zenn.dev
October 18, 2026 at 6:26 AM
今日のZennトレンド

Raspberry Pi 4でMCPサーバを構築する(3)
Raspberry Pi 4で、PythonライブラリPyODを用いた異常検知MCPサーバを構築した事例です。
Pico Wで収集したセンサデータを解析し、Mac mini上のLLMからツールとして呼び出す構成としています。
時系列データ向けのSpectral Residual手法を用いてCSVから異常点を検出し、LLMを通じてユーザーに回答する仕組みを実現しています。
Raspberry Pi 4でMCPサーバを構築する(3)
以前、Raspberry Pi Pico Wを使って、Ambientサービスのようなシステムを構築しました。https://zenn.dev/drbobt1122/articles/blog26082301https://zenn.dev/drbobt1122/articles/blog26082701また、Raspberry Pi 4(以下、Pi 4)をMCP(Model Context Pro
zenn.dev
October 14, 2026 at 12:14 AM
Printing and packing the last sticker orders from #pyod

Today the store officially closed. Thanks for #StickWithUs
September 21, 2026 at 11:52 AM
Six and a half years. 150+ designs. 7.500+ stickers on laptops around the world. 💜

We're closing PYOD.

Thank you for every order, every photo, every idea. Visit our site to order the free thank-you pack.

pimpyourowndevice.com

#stickers
September 2, 2026 at 6:45 AM
Alain Bashung - Bijou, Bijou
YouTube video by Sébastien MARTY
www.youtube.com
April 12, 2025 at 6:38 AM
Time to take down Dagan Gera!! (Jedi survivor Part 5)
www.twitch.tv/baatir_d_raps
www.youtube.com/watch?v=PyOd...
Twitch
Twitch is the world
www.twitch.tv
July 5, 2025 at 11:53 AM
My #PYOD stickers arrived today! I'm super happy with them! Now the important decision, where to put them on my laptop. Thanks @LuiseFreese @eliostruyf
January 1, 2025 at 10:57 AM
Gives me fuhnny feeling. Gotta find out if it’s true.
m.youtube.com/watch?v=pYOd...
Kylie Minogue - 09. Too Much
YouTube video by TheKylieMinogueCN
m.youtube.com
May 20, 2026 at 8:28 PM
We used to have PYOD.
December 14, 2024 at 5:37 AM
@wang.social You like python in the browser right ? Any chance you can push for CORS to be set on pypi.anaconda.org ? it's been blocking pyodide for 6+ months github.com/pyodide/micr... and github.com/pyodide/pyod...
Anaconda.org
pypi.anaconda.org
November 26, 2024 at 9:51 AM
Requiem For Evita
YouTube video by Andrew Lloyd Webber - Topic
www.youtube.com
November 13, 2025 at 9:07 AM
January 13, 2026 at 9:48 PM
I have a new work laptop and had to order some #PYOD stickers to appropriately decorate it.

https://pimpyourowndevice.com/
PimpYourOwnDevice - Dev & Tech Stickers
Do you think your laptop is a bit boring? We're here to provide you with unique dev and tech stickers to pimp up your own devices so that your laptop stands out.
pimpyourowndevice.com
January 12, 2025 at 10:35 AM
CVE-2026-15529 - yzhao062 pyod persistence.py pyod.utils.persistence.load deserialization
CVE ID : CVE-2026-15529

Published : July 13, 2026, 4:16 a.m. | 1 hour, 13 minutes ago

Description : A vulnerability was detected in yzhao062 pyod 3.5.0/3.5.1/3.5.2. Affected is the ...
CVE-2026-15529 - yzhao062 pyod persistence.py pyod.utils.persistence.load deserialization
A vulnerability was detected in yzhao062 pyod 3.5.0/3.5.1/3.5.2. Affected is the function pyod.utils.persistence.load of the file pyod/utils/persistence.py. Performing a manipulation of the argument path results in deserialization. The attack can be initiated remotely. The pull request to fix this issue requires some minor changes.
cvefeed.io
July 13, 2026 at 7:30 AM
Un #steelbook pour la nouvelle série Knuckles (prévue fin avril sur Paramount+) est déjà dispo en précommande sur la Fnac (visuel provisoire)

Fnac🇫🇷 ➡ edcollector.fr/2rdy
Fnac🇧🇪 ➡ edcollector.fr/o1xy

+d'info sur le blog➡ edcollector.fr/pyod
April 9, 2024 at 8:14 AM
My #PYOD stickers arrived today! I'm super happy with them! Now the important decision, where to put them on my laptop. Thanks @LuiseFreese @eliostruyf
December 30, 2024 at 8:32 PM
Oh goody, it might be fixed github.com/pyodide/pyod...
Fix pyproj by ryanking13 · Pull Request #436 · pyodide/pyodide-recipes
Close #356 I managed to make this all work locally, fingers crossed.
github.com
January 9, 2026 at 3:34 AM
Project Title: Customer transaction network analysis with pandas.

Description:Construct and analyze a dynamic customer–product transaction network using Pandas, NetworkX, TSFresh feature extraction, and PyOD anomaly detection. import pandas as pd import numpy as np import networkx as nx from…
Project Title: Customer transaction network analysis with pandas.
Description:Construct and analyze a dynamic customer–product transaction network using Pandas, NetworkX, TSFresh feature extraction, and PyOD anomaly detection. import pandas as pd import numpy as np import networkx as nx from tsfresh import extract_features from pyod.models.auto_encoder import AutoEncoder from sklearn.preprocessing import StandardScaler from datetime import datetime as dt # 1. Load & merge def load_data(trans_path: str, cust_path: str) -> pd.DataFrame: tx: pd.DataFrame = pd.read_csv(trans_path, parse_dates=['date']) cu: pd.DataFrame = pd.read_csv(cust_path) return tx.merge(cu, on='customer_id') # 2.
oluwadamilolaadegunwa.wordpress.com
May 23, 2025 at 5:32 PM
Project Title: Advanced knowledge graph construction and anomaly detection with pandas.

This project builds an Enterprise‑Scale Knowledge Graph by fusing tabular records, textual logs, and network data entirely in Pandas, augmented with graph analytics, NLP embeddings, temporal features, and…
Project Title: Advanced knowledge graph construction and anomaly detection with pandas.
This project builds an Enterprise‑Scale Knowledge Graph by fusing tabular records, textual logs, and network data entirely in Pandas, augmented with graph analytics, NLP embeddings, temporal features, and anomaly detection pipelines for actionable insights. The code leverages Dask for scalability, NetworkX and Faiss for graph operations, HuggingFace transformers for text, PyOD for outlier detection, and integrates real‑time streaming hooks. (
oluwadamilolaadegunwa.wordpress.com
May 23, 2025 at 7:41 AM
Lanzada en 2017, ha crecido hasta convertirse en la referencia estándar del sector, usada por empresas como Walmart, Databricks, IQVIA y Ericsson.
PyOD: la librería Python con 9,900 estrellas que detecta anomalías en tablas, audio, imágenes y más - Sinaptica
PyOD (Python Outlier Detection) es una de las librerías más completas para detección de anomalías en Python, con 9,900 estrellas en GitHub y más de 46 millones
sinapti.ca
July 20, 2026 at 10:00 PM