Đa số đội ngũ SaaS không dự báo doanh thu, họ chỉ dự báo "chuyển động" - những giai đoạn deal thay đổi, con số cập nhật, niềm tin tăng, nhưng lại khó giải thích **tại sao** quyết định mua hàng thay đổi. Dự báo lung lay không vì sai con số, mà do tiến độ được suy diễn, không được cam kết. Hãy chia sẻ deal gần đây mà ai cũng thấy tốt trong pipeline nhưng chẳng ai rõ lý do thúc đẩy! #SaaS #DoanhNghiep #QuanLyBánHang #Forecasting #KinhDoanh
#forecasting
He Predicted the Next Shutdown 3 Weeks Early
#math #statistics #forecasting
We're still clearing driveways, and now people are saying the news channels predict more snow this weekend. I checked NOAA. We've an estimated 3 our of 10 change of snow Saturday, 4 out of 10 Sunday. Well, less than 50 years ago I was considered qualified for a job as a statistician. Check me out. I figure the chance of no snow this weekend. is the product of the two days' chances: 0.6 x 0.7=0.42 That means around a 6 out of 10 chance that we will get snow.
Anybody care to correct this?
Then there's the psychology of it: if NBC says we're maybe, maybe not going to get snow, will they take a dusting for it? Do they actually have better prognosticators than the remaining NOAA staff? Or are people overstating what they hear?
#Apple #Weather and other #forecasting #apps tend to mislead users, especially ahead of big #storms like this weekend's. Find out why, and how to get more accurate information. https://www.cultofmac.com/news/how-weather-apps-mislead-users
I understand #weather #forecasting isn't supposed to be 100% accurate but I would think by this point, if a big set of storms is supposed to arrive in any area within 36 hours that I'd at least get a narrower gap in possibilities of snow total than .2 of an inch and 18 inches.
Research Flights Over the #Atlantic Could Help Improve #AtmosphericRiver #Forecasting
Global effort to better understand moisture-laden rivers in the sky. Scientists from the United States, Canada and Europe will soon start research flights over the #AtlanticOcean to learn more about atmospheric rivers in the hopes of giving battered residents of both #Europe and the West Coast of the #UnitedStates more time to prepare for the deluges they bring.
https://www.nytimes.com/2025/12/16/climate/atmospheric-rivers-forecasts.html
https://archive.ph/5ogdI
there is tarot reading and there is #btc #gold #forecasting #predictions reading more info:
AI predictions for 2026
Mantic Monday: The Monkey's Paw Curls
ARIMA vs. SARIMA in Python: Choosing the Right Model for Seasonal Data
When trend + noise is not enough, and seasonality must be modeled explicitly.
This post explains the difference, how to spot seasonality in real data, and how to evaluate both models correctly in Python (with clean time splits and practical checks).
#TimeSeries #Python #Forecasting #DataScience #ARIMA
@medium @towardsdatascience @programming @pythonclcoding @chartrdaily
Inside the Decades-Long Partnership Keeping Iconic Glacier National Park Road Safe [incl. video]
--
https://www.usgs.gov/programs/ecosystems-land-change-science-program/science/video-inside-decades-long-partnership <-- shared technical article / video
--
[I have been fortunate to ride and drive up and down this road, although not in the winter; a truly great experience and an impressive piece of road construction, especially as an engineering geologist]
#USGS #AvalancheForecasting #GlacierNationalPark #ScienceInAction #Transportation #risk #hazard #avalanche #SunRoad #GoingToTheSunRoad #GlacierNationalPark #infrastructure #fedscience #publicsafety #transportation #engineeringeology #forecasting #commerce #USA #Montana #landslide #massmovement #snow #climate #IntermountainWest
#USGS #NPS #USFS
Time Series Forecasting Analysis with Python
A practical workflow for finance data: clean the timeline, beat a baseline, and ship forecasts you can monitor.
This post walks through the real steps: missing dates, outliers, leakage-safe splits, baseline models, better models, and monitoring drift after deployment.
#Python #TimeSeries #Finance #DataScience #Forecasting
@chartrdaily @programming @pythonclcoding @theartificialintelligence @medium @towardsdatascience
Automate Change Point Detection in Time Series 📈
Change points in trends matter—a lot. Get them wrong, and your forecast doesn’t stand a chance.
In my latest post for The Forecaster, I show how to automate trend change point detection using piecewise regression and grid search, with full examples in R and Python.
🔗 https://theforecaster.substack.com/p/automatic-trend-change-point-detection
Deedy (@deedydas)
약 300개의 Kalshi 예측시장을 대상으로 현대 LLM들의 미래 예측 성능을 시험한 결과, Claude Opus 4.5가 가장 우수한 성과를 보였습니다. Claude Opus 4.5의 Brier score는 약 0.23으로 측정되었고, 인간 슈퍼포리케스터들의 범위(0.15–0.2)에 점차 근접하고 있으나 아직 차이가 존재합니다.
Want to learn how to model change points in time series trend? Created a tutorial about modeling change points with Piecewise Regression:
https://medium.com/@rami.krispin/piecewise-regression-for-time-series-forecasting-8cf53154a3fa?sk=3c9125c49233529de6d9c4072912fae7
We’re all at risk if Trump dismantles this legendary lab
Breaking up the National Center for Atmospheric Research would be a "genuinely shocking self-inflicted wound."
For more than half a century, the federally funded lab has been instrumental in the development of #weather models that have improved the #forecasting of extreme events like hurricanes, thus saving lives. Your go-to weather app can predict the future thanks in large part to the lab, also known as #NCAR.
https://grist.org/climate/were-all-at-risk-if-trump-dismantles-this-legendary-lab/
When Heavy Rain Meets #Glacial Landscapes – Record-Breaking #Flooding in #WashingtonState [USA]…
--
https://www.linkedin.com/posts/geological-society-of-america_geohazards-flooding-landslides-ugcPost-7405031818007187457-wuwy <-- shared GSA LinkedIn post
--
https://doi.org/10.1130/G39691.1 <-- shared paper
--
https://komonews.com/news/local/landslide-risk-western-washington-severe-flooding-heavy-rain-continues-for-dirt-soil-sediment-weather-forecast-winter-storm-pacific-northwest <-- shared media article
--
https://youtu.be/1yB34LibGo8?si=Z_Ug5gd3xLhvZGym <-- shared media video
--
#geohazards #Landslides #Geology #landslide #severeweather #landslides #landsliderisk #weather #westernwashington #washingtonflooding #flooding #winterconditions #storm #pnw #meteorology #science #massmovement #flood #extremeweather #fluvial #pluvial #rainfall #atmosphericriver #engineeringgeology #risk #hazard #mitigation #forecasting #riskmanagement #damage #infrastructure #cost #economics #publicsafety #PNW #PacificNorthwest #water #hydrology #glacial #sediment #sedimentology #stratigraphy #GIS #spatial #mapping #remotesensing #spatialanalysis #elevation #slopestability #froundwater #hydrology #model #flow #modeling #outwash #till #clay #watertable #lithology #landscape #geomorphology #geometry #USA
"...• Cutting NCAR isn’t a niche science fight — it touches nearly every corner of US weather and climate forecasting, from wildfire modeling to the computational backbone universities rely on...."
The Rise Of Community-Led Landslide Monitoring
Could an innovative natural-disaster monitoring system in Alaska be a model for the rest of the country?
--
https://seaklandslideworkinggroup.substack.com/p/article-in-sierra-about-southeast <-- shared technical #Substack
--
#communitybased #landslidemonitoring #NationalLandslidePreparednessAct #massmovement #engineeringgeology #risk #hazard #landslide #publicsafety #monitoring #Sitka #Alaska #debrisflow #risk #hazard #water #hydrology #pluvial #fluvial #rainfall #precipitation #AK #extremeweather #climatechange #atmosphericriver #coast #coastal #slope #steepness #Haines #Ketchikan #Wrangell #Juneau #propertydamage #cost #economics #soilmositure #sensors #raingauge #Skagway #emergencymanagement #forecasting #tool #online #dashboard #situationalawareness #NSF #grant #infrastructure #threshold #historic #analysis #geomorphology #geomorphometry #communication #buddysystem #risktolerance #SitkaTribeofAlaska
#NationalScienceFoundation #USFS
Kicking off my new newsletter about #forecasting with an introduction to piecewise regression for modeling changes in the trend 👇🏼
https://theforecaster.substack.com/p/piecewise-regression-for-time-series
Code is available both in #Rstats and #Python
Subscribe to get an email notification.
Client Info
Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst