[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.] #  @predict_addict Valeriy M., PhD, MBA, CQF Valeriy M., PhD, MBA, CQF posts on X about realworld, prediction, french polynesia, kazakhstan the most. They currently have XXXXXX followers and XX posts still getting attention that total XXXXX engagements in the last XX hours. ### Engagements: XXXXX [#](/creator/twitter::1464965267830804482/interactions)  - X Week XXXXXX +132% - X Month XXXXXXX -XX% - X Months XXXXXXXXX -XX% - X Year XXXXXXXXX -XX% ### Mentions: XX [#](/creator/twitter::1464965267830804482/posts_active)  - X Week XX -XX% - X Month XXX -XXXX% - X Months XXX +313% - X Year XXX +83% ### Followers: XXXXXX [#](/creator/twitter::1464965267830804482/followers)  - X Week XXXXXX +0.19% - X Month XXXXXX +1.50% - X Months XXXXXX +11% - X Year XXXXXX +51% ### CreatorRank: XXXXXXX [#](/creator/twitter::1464965267830804482/influencer_rank)  ### Social Influence [#](/creator/twitter::1464965267830804482/influence) --- **Social category influence** [countries](/list/countries) XXXX% [stocks](/list/stocks) XXXX% [finance](/list/finance) XXXX% **Social topic influence** [realworld](/topic/realworld) 4.92%, [prediction](/topic/prediction) 3.28%, [french polynesia](/topic/french-polynesia) 3.28%, [kazakhstan](/topic/kazakhstan) 3.28%, [luxury](/topic/luxury) 1.64%, [gps](/topic/gps) 1.64%, [supreme](/topic/supreme) 1.64%, [applied](/topic/applied) 1.64%, [faster](/topic/faster) 1.64%, [matrix](/topic/matrix) XXXX% **Top accounts mentioned or mentioned by** [@jlhrzn](/creator/undefined) [@unknownanony27](/creator/undefined) [@timsucksatchess](/creator/undefined) [@paddy_mullen](/creator/undefined) [@darthjohannes](/creator/undefined) [@mattiaslamotte](/creator/undefined) [@artea1_](/creator/undefined) [@pl_ankit](/creator/undefined) ### Top Social Posts [#](/creator/twitter::1464965267830804482/posts) --- Top posts by engagements in the last XX hours "Translation: Stability isnt a luxury; its a profit driver. SDC in X Simple Steps: ✅ Track: Generate sequential forecasts (e.g. daily updates). ✅ Measure: Calculate SDC by comparing each forecast to the prior one. (Works for single-point OR probabilistic forecasts) Act:"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1947991068462420417) 2025-07-23 12:03:57 UTC 34.1K followers, XX engagements "There have been scalable approximations proposed since the Stone Age (Sparse GPs inducing points variational fluff) but they all compromise accuracy and rarely work in the wild. Deep learning eats GPs for breakfast in both performance and runtime"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1947232134038471118) 2025-07-21 09:48:13 UTC 34.1K followers, XXX engagements "✔ Chasing peak performance with unlimited resources Deep ensembles can edge out CatBoostbut at a high cost. 🔬 Final Verdict Still listening to Kaggle fairytales 🤦 ✅The data doesnt lie: CatBoost reigns supreme for most real-world tabular tasks. Want the absolute best"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1946965731041566900) 2025-07-20 16:09:37 UTC 34.1K followers, XX engagements "An interesting rant by a student on Reddit and totally valid. Forecasting is one of the most in-demand industry skills yet time series classes in academia barely scratch the surface. I had the same realization while studying machine learning at UCLarguably the top ML program in Europe"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1947688439790350507) 2025-07-22 16:01:25 UTC 34.1K followers, 4579 engagements "🚨 Next Cohort Starts This September Join here The Applied Conformal Prediction course is backand attracting an even stronger global cohort. From Bayesian skeptics to quant researchers from PhD students to VPs of Engineering this is where real-world ML pros go to master uncertainty"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1946608657728168091) 2025-07-19 16:30:44 UTC 34.1K followers, 1036 engagements "🚀 TorchCP is quickly becoming the go-to library for conformal prediction modern fast and built for real-world machine learning. 🔍 TorchCP stands out by: Unifying conformal methods for classification & regression"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1945878705026351334) 2025-07-17 16:10:10 UTC 34.1K followers, XXX engagements "Result 🚨 1.4GB bloated to the max with float64 int64 and object types. A few lines of code shrinks it by 5x: sales_df: 456MB 96MB prices_df: 957MB 85MB Faster loads. Lower RAM. Happier kernels"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1946937041721123041) 2025-07-20 14:15:37 UTC 34.1K followers, XXX engagements "These estimates are conditional on your prior and kernel choices which are totally arbitrary and usually unjustified. And worseTHEY DONT EVEN PROVIDE COVERAGE. Your GP can be XX% confident and still be catastrophically wrong. Calibration Garbage. Reliability Nonexistent"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1947232090220609542) 2025-07-21 09:48:02 UTC 34.1K followers, XXX engagements "Youre better off flipping a coin. 🐢 Slower Than Evolution GPs choke on data sizes over 10k points. Why Because youre inverting an $n times n$ matrix every time like its 1953"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1947232112211317075) 2025-07-21 09:48:08 UTC 34.1K followers, XXX engagements "Looking ahead to future chapters of my book '📘 Mastering Modern Time Series Forecasting ' what additional topics would you like to see coveredbeyond those already in progress (ML DL Transformers and Foundational Models)"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1948072903141900739) 2025-07-23 17:29:08 UTC 34.1K followers, 1893 engagements "Forecasting Fever Goes Global 🌍📈 Its one thing to see interest from Germany or Canada but when Mauritius Kazakhstan and French Polynesia start buying your book on forecasting you know the future is calling from every corner of the planet"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1945552294084231534) 2025-07-16 18:33:08 UTC 34.1K followers, XXX engagements "📈 JP Morgan and top quant firms are using computer vision to trade stocks. Why arent you Wall Street isn't just looking at candlesticks anymore it's scanning time series as images"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1946143428489736561) 2025-07-18 09:42:05 UTC 34.1K followers, 22.8K engagements "As I was evaluating the stack for a new chapter I revisited a few libraries. I remembered functimea few years ago it made bold claims upon its release. Initially it touted itself as the most powerful forecasting library"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1947014504895610985) 2025-07-20 19:23:26 UTC 34.1K followers, XXX engagements "💡 Pro tip: Always run optimize_dataframe() or use dtype_diet before anything else. Stop grinding your code to a halt for no reason. Compressing data is good data hygiene. #timeseries #forecasting #data"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1946937063535702344) 2025-07-20 14:15:42 UTC 34.1K followers, XXX engagements "Combine tuned modelsincluding deep nets and foundation methodswith CatBoost as your backbone. The moral Stop following folklore 🧚. Start trusting evidence. Great work by authors Much needed Arena. My new book Mastering CatBoost is now available in Early Release"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1946965752822599742) 2025-07-20 16:09:43 UTC 34.1K followers, XX engagements "Theres still a hefty backlog of chapters waiting to be reviewed and the best way to keep the dopamine flowing is to jump in as a reviewer"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1945499171818979479) 2025-07-16 15:02:02 UTC 34.1K followers, XX engagements "🧙♂ Priors Kernels and Unicorns Ah yes kernel choice: the arcane spellbook of GP practitioners. You basically guess a kernel cross your fingers and pray to the Gaussian gods. Rational Quadratic Matern Dot-Product You might as well use horoscopes"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1947232155844637008) 2025-07-21 09:48:18 UTC 34.1K followers, XXX engagements "Low SDC = High Stability: Fewer plan changes lower costs smoother ops. High SDC = High Volatility: Constant fire drills soaring operational costs. Is SDC Your Missing Metric Ask Yourself: ✅ Do plan adjustments burn cash and time (Switching costs)"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1947991090751004769) 2025-07-23 12:04:02 UTC 34.1K followers, XX engagements "✨ Kazakhstan Central Asias rising giant leveraging forecasting to shape policy trade and oil markets. ✨ French Polynesia Remote paradise meets global thinking. Even paradise wants to anticipate whats next"  [@predict_addict](/creator/x/predict_addict) on [X](/post/tweet/1945552338049008070) 2025-07-16 18:33:18 UTC 34.1K followers, XX engagements
[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]
@predict_addict Valeriy M., PhD, MBA, CQF
Valeriy M., PhD, MBA, CQF posts on X about realworld, prediction, french polynesia, kazakhstan the most. They currently have XXXXXX followers and XX posts still getting attention that total XXXXX engagements in the last XX hours.
Engagements: XXXXX #
- X Week XXXXXX +132%
- X Month XXXXXXX -XX%
- X Months XXXXXXXXX -XX%
- X Year XXXXXXXXX -XX%
Mentions: XX #
- X Week XX -XX%
- X Month XXX -XXXX%
- X Months XXX +313%
- X Year XXX +83%
Followers: XXXXXX #
- X Week XXXXXX +0.19%
- X Month XXXXXX +1.50%
- X Months XXXXXX +11%
- X Year XXXXXX +51%
CreatorRank: XXXXXXX #
Social Influence #
Social category influence countries XXXX% stocks XXXX% finance XXXX%
Social topic influence realworld 4.92%, prediction 3.28%, french polynesia 3.28%, kazakhstan 3.28%, luxury 1.64%, gps 1.64%, supreme 1.64%, applied 1.64%, faster 1.64%, matrix XXXX%
Top accounts mentioned or mentioned by @jlhrzn @unknownanony27 @timsucksatchess @paddy_mullen @darthjohannes @mattiaslamotte @artea1_ @pl_ankit
Top Social Posts #
Top posts by engagements in the last XX hours
"Translation: Stability isnt a luxury; its a profit driver. SDC in X Simple Steps: ✅ Track: Generate sequential forecasts (e.g. daily updates). ✅ Measure: Calculate SDC by comparing each forecast to the prior one. (Works for single-point OR probabilistic forecasts) Act:"
@predict_addict on X 2025-07-23 12:03:57 UTC 34.1K followers, XX engagements
"There have been scalable approximations proposed since the Stone Age (Sparse GPs inducing points variational fluff) but they all compromise accuracy and rarely work in the wild. Deep learning eats GPs for breakfast in both performance and runtime" @predict_addict on X 2025-07-21 09:48:13 UTC 34.1K followers, XXX engagements
"✔ Chasing peak performance with unlimited resources Deep ensembles can edge out CatBoostbut at a high cost. 🔬 Final Verdict Still listening to Kaggle fairytales 🤦 ✅The data doesnt lie: CatBoost reigns supreme for most real-world tabular tasks. Want the absolute best" @predict_addict on X 2025-07-20 16:09:37 UTC 34.1K followers, XX engagements
"An interesting rant by a student on Reddit and totally valid. Forecasting is one of the most in-demand industry skills yet time series classes in academia barely scratch the surface. I had the same realization while studying machine learning at UCLarguably the top ML program in Europe" @predict_addict on X 2025-07-22 16:01:25 UTC 34.1K followers, 4579 engagements
"🚨 Next Cohort Starts This September Join here The Applied Conformal Prediction course is backand attracting an even stronger global cohort. From Bayesian skeptics to quant researchers from PhD students to VPs of Engineering this is where real-world ML pros go to master uncertainty" @predict_addict on X 2025-07-19 16:30:44 UTC 34.1K followers, 1036 engagements
"🚀 TorchCP is quickly becoming the go-to library for conformal prediction modern fast and built for real-world machine learning. 🔍 TorchCP stands out by: Unifying conformal methods for classification & regression" @predict_addict on X 2025-07-17 16:10:10 UTC 34.1K followers, XXX engagements
"Result 🚨 1.4GB bloated to the max with float64 int64 and object types. A few lines of code shrinks it by 5x: sales_df: 456MB 96MB prices_df: 957MB 85MB Faster loads. Lower RAM. Happier kernels" @predict_addict on X 2025-07-20 14:15:37 UTC 34.1K followers, XXX engagements
"These estimates are conditional on your prior and kernel choices which are totally arbitrary and usually unjustified. And worseTHEY DONT EVEN PROVIDE COVERAGE. Your GP can be XX% confident and still be catastrophically wrong. Calibration Garbage. Reliability Nonexistent" @predict_addict on X 2025-07-21 09:48:02 UTC 34.1K followers, XXX engagements
"Youre better off flipping a coin. 🐢 Slower Than Evolution GPs choke on data sizes over 10k points. Why Because youre inverting an $n times n$ matrix every time like its 1953" @predict_addict on X 2025-07-21 09:48:08 UTC 34.1K followers, XXX engagements
"Looking ahead to future chapters of my book '📘 Mastering Modern Time Series Forecasting ' what additional topics would you like to see coveredbeyond those already in progress (ML DL Transformers and Foundational Models)" @predict_addict on X 2025-07-23 17:29:08 UTC 34.1K followers, 1893 engagements
"Forecasting Fever Goes Global 🌍📈 Its one thing to see interest from Germany or Canada but when Mauritius Kazakhstan and French Polynesia start buying your book on forecasting you know the future is calling from every corner of the planet" @predict_addict on X 2025-07-16 18:33:08 UTC 34.1K followers, XXX engagements
"📈 JP Morgan and top quant firms are using computer vision to trade stocks. Why arent you Wall Street isn't just looking at candlesticks anymore it's scanning time series as images" @predict_addict on X 2025-07-18 09:42:05 UTC 34.1K followers, 22.8K engagements
"As I was evaluating the stack for a new chapter I revisited a few libraries. I remembered functimea few years ago it made bold claims upon its release. Initially it touted itself as the most powerful forecasting library" @predict_addict on X 2025-07-20 19:23:26 UTC 34.1K followers, XXX engagements
"💡 Pro tip: Always run optimize_dataframe() or use dtype_diet before anything else. Stop grinding your code to a halt for no reason. Compressing data is good data hygiene. #timeseries #forecasting #data" @predict_addict on X 2025-07-20 14:15:42 UTC 34.1K followers, XXX engagements
"Combine tuned modelsincluding deep nets and foundation methodswith CatBoost as your backbone. The moral Stop following folklore 🧚. Start trusting evidence. Great work by authors Much needed Arena. My new book Mastering CatBoost is now available in Early Release" @predict_addict on X 2025-07-20 16:09:43 UTC 34.1K followers, XX engagements
"Theres still a hefty backlog of chapters waiting to be reviewed and the best way to keep the dopamine flowing is to jump in as a reviewer" @predict_addict on X 2025-07-16 15:02:02 UTC 34.1K followers, XX engagements
"🧙♂ Priors Kernels and Unicorns Ah yes kernel choice: the arcane spellbook of GP practitioners. You basically guess a kernel cross your fingers and pray to the Gaussian gods. Rational Quadratic Matern Dot-Product You might as well use horoscopes" @predict_addict on X 2025-07-21 09:48:18 UTC 34.1K followers, XXX engagements
"Low SDC = High Stability: Fewer plan changes lower costs smoother ops. High SDC = High Volatility: Constant fire drills soaring operational costs. Is SDC Your Missing Metric Ask Yourself: ✅ Do plan adjustments burn cash and time (Switching costs)" @predict_addict on X 2025-07-23 12:04:02 UTC 34.1K followers, XX engagements
"✨ Kazakhstan Central Asias rising giant leveraging forecasting to shape policy trade and oil markets. ✨ French Polynesia Remote paradise meets global thinking. Even paradise wants to anticipate whats next" @predict_addict on X 2025-07-16 18:33:18 UTC 34.1K followers, XX engagements