Python 時系列分析 1,000本ノック
– ノック87: ヨー＝ジョンソン変換 –
salesanalytics.co.jp/knocks/…
#時系列分析 #時系列予測 #YeoJohnson #BoxCox
1
5
412
Mar 15
🔥 AutoQuant: Pro-grade R AutoML package that automates the entire ML lifecycle … data prep → modeling → tuning → evaluation → interpretation.
Core Features:
✅ 20 algorithms: CatBoost, LightGBM, XGBoost, H2O (AutoML, GBM, DL), Ranger, glmnet
✅ GPU acceleration across models
✅ Multi-armed bandit grid tuning (no more manual grids)
✅ Auto feature engineering: BoxCox, YeoJohnson, Asinh, quantile binning, interactions, lags
✅ Time series forecasting with calendars, lags, rolling stats
✅ Model evaluation: Confusion matrices, ROC, calibration plots, feature importance
✅ Interpretation: SHAP, ICE, LIME plots PDF reports with thresholds/actions
✅ Big data optimized (data.table backend, parallel processing)
Unique strengths:
• Production-ready (handles 100GB datasets)
• Handles multi-type targets (numeric, binary, categorical, count)
• Automatic train/test splitting with time-series awareness
• Cross validation ensemble stacking
• Leaderboard scoring with statistical tests
Beats most AutoML tools for serious ML work. No Python dependency hell … just R.
github.com/AdrianAntico/Auto… #Rstats #MachineLearning #AutoML
7
12
125
19 Oct 2025
Dev log
> Learnt mathematical transformation
Log,reciprocal,power,boxcox etc
> Done with day 1 and 2 of #TUFWinterArc by @takeUforward_
Cause I don't know that this arc or program is going on
>Doing DSA in cpp
6
123
12 Mar 2024
أغلب خوارزميات #علم_البيانات و #الذكاء_الاصطناعي تفترض ان العينة جات من توزيع بيانات طبيعي " Normal distribution " وفي الواقع غالبًا البيانات الحقيقة ماتكون تحت هذا التوزيع❗️
طيب كيف نحل هالمشكلة ؟
هالمشكلة ممكن تنحل بخطوة بسيطة :
نستخدم (boxcox) الي موجود بمكتبة scipy، وراح يحول البيانات للتوزيع الطبيعي ✅💡
1
28
15,613
🚀Day 63 . The road to mastering #MachineLearning
I studied how to transform & standardize data using Scikit-Learn and numpy using techniques such as
-Log, Reciprocal, exponential, square root & BoxCox transformations
-RobustScaler
-MinMax Scaler
-Standard Scaler
#DataScientist
1
1
306
18 Feb 2024
心の底から美しいと思う曲。
この曲が沢山の人に聞かれないなんて本当におかしいよ。
【PV】SUN TIGER／福岡史朗＆BOXCOX youtu.be/hMFTBcAczFY?si=bVTx… @YouTubeより
2
7
776
4 Feb 2023
then asking you how to code boxcox plot of the relationship between x and z
you: my bf plays league of legends 🤓🤓
1
61
23 Dec 2022
Day 33 of #66DaysOfData R3
- Learned about how to use the .apply() function to iterate through the DataFrame and how it is faster than .iterrows()!
- Learned that boxcox() function in scipy.stats is great to transform skewed data…lnkd.in/g8SGNNXH lnkd.in/g2T-aeBq
4
66
22 Oct 2022
How to resolve Inf. $y values when using MASS::boxcox? #tidyverse #rstats community.rstudio.com/t/how-…
2
1
27 Sep 2022
the fact that a guy with the last name cox teamed up with a guy with the last name box to come up with a data transformation and then named it boxcox... like holy fuck. camp.
1
1
29 Jan 2022
#66daysofdata R2 D20: Really chill day, nearly done nothing. But I did more than 5 minutes :)
- Read about transformation - boxcox, power transformer, yeo-johnson - #sklearn has great documentation
- applied these on song popularity data
- 5 minutes of #mysql and @tableau
19
Introducing the new add-in for @IHSEViews: BOXCOX. This add-in can apply power transformations to the series of interest and provides alternative methods to estimate the optimal lambda parameter to be used in transformation:
#EViews #EViewsGuide #BOXCOX bit.ly/3cFnSoF
2
13 Jun 2021
例えば一年前のM5なんかはコンペとしてはシェイク祭りでアレだったけど、poisson分布gamma分布、指数平滑による予測とかARIMA系も試したし、boxcoxとかも触れて、初めてだったことを差し引いても得るものがすごく多かった覚えがある
3
29 Apr 2021
Normalize all your variables using boxcox and run two models, only with those continuous and other only with categorical. Use a mix of both, with all signficant p-values.
3
