- Видео 54
- Просмотров 515 615
paretos
Германия
Добавлен 7 янв 2012
Paretos is the leading AI-based decision intelligence platform to make effective, data-driven decisions across entire organizations, enabling business users to evaluate complex data, predict future scenarios and take optimal actions via a no-code UI and integrations with no prior data science knowledge needed. On this channel we want to give decision makers a better overview of technology and applications in order to create real business value and minimising sunken costs of IT projects.
For our Geeks and Engineers: We focus on explaining theoretical knowledge about AI, math, and optimisation and how to bring technology to application (as this is often the hard part). We want to explain these topics easily and understandably and give you easy access to potential applications for these approaches. Missing a video on a certain topic? Reach out and we'll see what we can do.
For our Geeks and Engineers: We focus on explaining theoretical knowledge about AI, math, and optimisation and how to bring technology to application (as this is often the hard part). We want to explain these topics easily and understandably and give you easy access to potential applications for these approaches. Missing a video on a certain topic? Reach out and we'll see what we can do.
Hot tips for cold starts - how to improve forecasting accuracy for (new) cold start products
Welcome to AI Forecating Academy - powered by paretos! Session from December 12, 2023 in Heidelberg with Daria Mokrytska (AI Scientist - paretos): Hot tips for cold starts - how to improve forecasting accuracy for (new) cold start products
Daria Mokrytska is an AI Scientist at paretos in Heidelberg, Germany, where the central focus of her work is within Timeseries Forecasting, with a specific emphasis on forecasting demand. Before her current role, she gained experience as a Research Assistant at the Heidelberg Institute for Theoretical Studies, focusing on astrophysics and data analysis. Daria holds a Master’s degree in Physics from the National University of Kharkiv in Ukraine.
Daria Mokrytska is an AI Scientist at paretos in Heidelberg, Germany, where the central focus of her work is within Timeseries Forecasting, with a specific emphasis on forecasting demand. Before her current role, she gained experience as a Research Assistant at the Heidelberg Institute for Theoretical Studies, focusing on astrophysics and data analysis. Daria holds a Master’s degree in Physics from the National University of Kharkiv in Ukraine.
Просмотров: 331
Видео
Propabilistic Timeseries Forecasting - Introduction to Autogluon with founder Oleksandr Shchur
Просмотров 1,1 тыс.10 месяцев назад
Welcome to AI Forecating Academy - powered by paretos! Session from December 12, 2023 in Heidelberg with Oleksandr Shchur (Applied Scientist - AWS AI): AutoGluon-TimeSeries: AutoML for Probabilistic Time Series Forecasting. Oleksandr Shchur is an Applied Scientist at Amazon Web Services, where he works on time series forecasting. He is a core contributor to AutoGluon (auto.gluon.ai/) - an open-...
How to win M6 Financial Time Series Forecasting Competition with Moretza Khani
Просмотров 47510 месяцев назад
Welcome to AI Forecating Academy - powered by paretos! Session from December 12, 2023 in Heidelberg with Morteza Khani (Analytics Consultant - iqast): Leveraging Behavioral Science Insights to Triumph in M6 Financial Time Series Forecasting Competition. Morteza is a leading expert in time series analysis and AI, known for his innovative approach in data-driven forecasting. Representing team #ma...
trade/off Summit 2023 Aftermovie - Business Growth in the Era of AI
Просмотров 211Год назад
Der trade/off Summit - die Konferenz für Business Growth und Innovation im KI-Zeitalter. Zusammen mit führenden Köpfen und Visionären aus Industrie, Digitalwirtschaft und Wissenschaft beleuchten wir die Potentiale moderner Technologien für den wirtschaftlichen Erfolg. Die Konferenz bietet ein inspirierendes Forum für Austausch und Weiterbildung rund um die Frage, wie Professionals und Entscheid...
AI als Risikomanager: Ist das Supply Chain Management für die nächste Krise gerüstet?
Просмотров 146Год назад
Angesichts wachsender globaler Unsicherheiten und steigenden Marktanforderungen steht die Resilienz der Lieferketten auf dem Prüfstand. In diesem Panel diskutieren führende Expert:innen darüber, inwiefern Künstliche Intelligenz ein Schlüsselwerkzeug ist, um Komplexitäten in Krisenzeiten zu bewältigen, welche KI-gestützten Strategien für die Optimierung von Lieferketten wirklich Früchte tragen u...
Wie man in Extremsituationen Ressourcen und Unsicherheiten meistert: Learnings vom Mount Everest
Просмотров 45Год назад
Das Bergsteiger-Credo "Climb high, sleep low" steht für den klugen Umgang mit Ressourcen in kritischen Phasen - eine Maxime, die angesichts der aktuellen Wirtschaftslage relevanter ist denn je. Die Bergsteigerin Helga Hengge, die als erste deutsche Frau den Mount Everest und die Seven Summits bezwungen hat, zieht in einem fesselnden Vortrag Vergleiche zwischen den Höhen und Tiefen ihrer Expedit...
Culture eats data-driven decision making for breakfast - Panel Discussion
Просмотров 54Год назад
Damit Unternehmen das Potenzial moderner Technologien voll ausschöpfen können, ist eine fortschrittsbereite Unternehmenskultur unerlässlich. Doch wie gelingt der Wandel zur datengetriebenen Organisation? Die Panel-Teilnehmer diskutieren, vor welchen Herausforderungen Unternehmen heute stehen, welche Maßnahmen wirklich vielversprechend sind und was Mitarbeitende jetzt von ihren Führungskräften e...
How AI helps to manage risks in the energy transition
Просмотров 65Год назад
Der EU AI Act stuft viele KI-Anwendungen in kritischen Infrastrukturen als hochriskant ein - und doch bietet Künstliche Intelligenz für Stromversorgungsunternehmen wie die Elia Group enorme Chancen. Rachel Berryman, Stellvertretende Leiterin des AI Center of Excellence, zeigt auf, wie KI dazu beiträgt, Stromversorgungssysteme sicherer und kosteneffizienter zu gestalten und geht auf den steigend...
Navigating Tomorrow: Decision Intelligence als Erfolgsfaktor für die zukunftsfähige Organisation
Просмотров 92Год назад
Daten allein führen nicht zum Erfolg; es kommt darauf an, wie man sie nutzt. Heiko Kahrels, Partner für Supply Chain & Operations bei EY, beleuchtet die essentielle Rolle von Decision Intelligence (DI) in modernen Unternehmen. Er beschreibt den Weg einer nachhaltigen DI-Transformation und erklärt, wie Daten, Analytik und Technologie zusammenwirken, um Entscheidungsprozesse zu optimieren, damit ...
KI-Regulierung vs. Innovationskraft: Wo steht Europa?
Просмотров 85Год назад
Mit dem Fokus auf den EU AI Act nimmt dieses Panel die europäische KI-Regulierung kritisch unter die Lupe. Wie groß ist die aktuelle Wissenslücke bei Entscheidungsträgern und in der breiten Öffentlichkeit und was muss man wirklich über die KI-Regulierung wissen, um fundierte Entscheidungen zu treffen? Unsere Expert:innen diskutieren, ob die europäischen Ansätze im globalen Vergleich standhalten...
Human is the next big thing: Was wir durch KI über unsere Fähigkeiten und Stärken lernen
Просмотров 170Год назад
“Die wichtigsten Innovationen im Zeitalter Künstlicher Intelligenz sind Mensch und Menschlichkeit.” Mit diesem Statement lädt uns Dr. Rebekka Reinhard zu einer Reflexion über unsere Rolle im KI-Zeitalter ein. Sie motiviert uns, die mächtige Technologie als Erweiterung und Bereicherung unserer menschlichen Fähigkeiten zu verstehen und fordert zu ethischem Handeln sowie kritischem Denken auf, um ...
Simplicity is complicated: 6 Schritte zur Vermeidung einer KI-Katastrophe von Sam Edds
Просмотров 26Год назад
Viele Unternehmen, die KI entwickeln, stoßen auf unbeabsichtigte Nebeneffekte ihrer Systeme. Dort, wo Entscheidungen über Menschen getroffen werden - zum Beispiel bei der Polizeiarbeit oder bei Kreditvergaben - sind die Risiken moderner Algorithmen besonders heikel. Sam Edds, Datenpionierin bei Yelp und HP, präsentiert praktische Ansätze, um solchen Risiken zu begegnen und eine sichere und zuve...
Lufthansa Innovation Hub: Pioneering Tomorrow's Travel mithilfe von Künstlicher Intelligenz
Просмотров 153Год назад
Schon immer hat die Lufthansa den Kurs der Luftfahrtbranche maßgeblich geprägt. Heute steuert Christine Wang, Innovationschefin des Lufthansa Innovation Hubs, das Unternehmen in die Ära der Künstlichen Intelligenz. Im Gespräch teilt die Top-Managerin Einblicke in die innovativen Mobilitätsansätze und Projekte des Hubs, skizziert ihre Vision einer nahtlosen Travel Experience und beleuchtet die s...
How to achieve Supply Chain Excellence with Forecasting and Inventory Optimization
Просмотров 203Год назад
How to achieve Supply Chain Excellence with Forecasting and Inventory Optimization
Decision Intelligence: Wie KI die Prognoseprozesse der Otto Group revolutioniert
Просмотров 640Год назад
Decision Intelligence: Wie KI die Prognoseprozesse der Otto Group revolutioniert
Beyond the Hype: Was wir 2023 wirklich über KI lernen müssen. Insights aus dem Silicon Valley
Просмотров 525Год назад
Beyond the Hype: Was wir 2023 wirklich über KI lernen müssen. Insights aus dem Silicon Valley
Web Summit 2023: You don’t want an LLM managing your Supply Chain
Просмотров 687Год назад
Web Summit 2023: You don’t want an LLM managing your Supply Chain
Beyond Toy Datasets: Timeseries Forecasting for Real Business Problems (PyData Südwest)
Просмотров 284Год назад
Beyond Toy Datasets: Timeseries Forecasting for Real Business Problems (PyData Südwest)
trade/off - How do companies create a more efficient decision-making culture?
Просмотров 3392 года назад
trade/off - How do companies create a more efficient decision-making culture?
trade/off - Is online retail sustainable without automated decision-making processes?
Просмотров 1152 года назад
trade/off - Is online retail sustainable without automated decision-making processes?
trade/off - Doppelgänger Tech Talk Live Podcast
Просмотров 6062 года назад
trade/off - Doppelgänger Tech Talk Live Podcast
trade/off - Controlled thrills: How calculable is the risk in freestyle kayaking?
Просмотров 1222 года назад
trade/off - Controlled thrills: How calculable is the risk in freestyle kayaking?
trade/off - Machine Intelligence: Is Decision Intelligence the New Data Science?
Просмотров 2282 года назад
trade/off - Machine Intelligence: Is Decision Intelligence the New Data Science?
trade/off - AI in the mobility industry: digitization in DB's rail production
Просмотров 592 года назад
trade/off - AI in the mobility industry: digitization in DB's rail production
trade/off - Can AI help "neutralize" judicial decision making?
Просмотров 1092 года назад
trade/off - Can AI help "neutralize" judicial decision making?
trade/off - Opportunities and potentials of data-driven decisions for the logistics industry
Просмотров 1332 года назад
trade/off - Opportunities and potentials of data-driven decisions for the logistics industry
trade/off - Digital Trust: Which factors influence entrepreneurial decision-making?
Просмотров 682 года назад
trade/off - Digital Trust: Which factors influence entrepreneurial decision-making?
trade/off - How companies use AI-based predictions to make complex decisions
Просмотров 2272 года назад
trade/off - How companies use AI-based predictions to make complex decisions
trade/off - Decisions in Uncertainty and the Courage to Take Risks: Poker World Champion Learnings
Просмотров 872 года назад
trade/off - Decisions in Uncertainty and the Courage to Take Risks: Poker World Champion Learnings
Please help. I created a group, made it public, and tried to add @RawDataBot. It gets added but immediately leaves the group. (( How can I fix this? I need this bot to stay in the group and immediately add JSON to it.
@26:30 So, does the presenter say that ensambling different models can somehow smooth out the need for feature engineering?
Is this about Knots and Splines?
Tutorial en castellano de optimizacion bayesiana, por si a alguien le interesa: ruclips.net/video/nNRGOfneMdA/видео.html
2:36 Sorry but what most people, including you, fail to explain is HOW DO N-DIMENSIONAL NORMAL DISTRIBUTIONS LEAD TO GAUßIAN PROCESSES? Are the (infinite) Gaußian distributions added on top of each other on the x-axis, on the y-axis? How can we understand this visually? This is such an abstract concept that not even this "introduction" manages to explain THE SLIGHTEST. This is not an introduction, it's just another blabla to trick the viewer into understanding a concept that can't be understood if it isn't explained better. To get more customers.
Thank you. I have some questions about how I can change your code to be fitted for my model. My objective functions are min and max, so how can I define it. Also, I didn't know how to define constraints. How can I do that?
Thanks. Other videos?
excellent explanation with visual intuition. One thing that was not clear to me is what differentiates minimization and maximization problems. For example, let's say my f_objective returns the metric R2 (maximization), how do I configure the search for this? and if I change the metric to mean squared error (MSE, minimization), what changes in the optimization???
hello,a perfect video.can you share you code?"I want to continue following your approach and proceed further."
J'ai rien compris.
That's cool. can you explain the NSGA2 with a csv file? it would be of great help. Thanks in advance
Great video, thank you!
Can you suggest how to do GPR with poisson likelihood? Should i use approximation for inference like using laplace approximation?
after hours trying to find that chat id using telegram bots I found you and I finally found the correct chat id. thanks a lot man!
Thanks man!! Very well explained
thank you
how can we get the entire code of this video? thanks
It should be available in the repo. If not ping us again so we will check.
Loved the vibe here guys! Thanks for an awesome video!
You are literally reading a script on the video, bro
Thank you so much!
How do you find Pareto line ?
There are eather libraries like pymoo or you can implement it by yourself as we show in the NSGA-2 Video.
Thank you for the perfect explanation. Can we call the last part as bayesian optimization, i.e. the combination of gaussian process and conditional probability mechanism?
Definetly!
that slack notification sound at 4:30 got me checking my slack 👀
🤣🤣🤣🤣🤣🤣
Thank you so much! it has been helpful
00:02 NSGA-II helps find optimal solutions for multi-objective optimization problems. 02:34 Genetic algorithm uses survival of the fittest principle to find optimal solutions. 05:10 NSGA-II uses non-dominated sorting for population selection 07:49 Dominated sorting identifies individuals that dominate others 10:29 NSGA-II optimization involves finding different fronts based on domination count. 13:21 Non-dominated sorting results in sorted population RT due to its performance on target indicators. 16:04 NSGA-II optimization involves parent population selection and offspring creation 18:30 NSGA-II uses crossover and mutations in iterative process for optimization. Crafted by Merlin AI.
Video style looks copy of 3blue1brown channel🤣🤣
The library he build to do such animations is just great. We could not resist using it 🚀
Excellent info. It works perfect. Thx a lot.
کیا آپ مہندس ھو؟ 😮
کمال کی ویڈیو ❤
Thanks!!!!
good explanation, thanks.
This is the first video I've watched about NGSA-II and also the last I need.🤩
This was the intention! We are now getting on new content for other optimization types as the combination of black box and linear optimization is powerful. Stay tuned
Hello sir I run the code and faced to this problem please help me. #ModuleNotFoundError: No module named 'pymoo.model'
Have you found the solution yet?
Did you check the newest pymoo version? I think there is a change in the inferface. The video was done on version 0.4.0
Hi! Könntet ihr mal was zum Thema KI Software machen?
Klar. Könntest du etwas spezifischer werden was dich interessieren würde?
非常感谢,讲解知识简单易懂,大师
Still useful today. Thank you.
Good video.
Can you plz make a video that how to take data from excel sheet and use it to optimize model
We normally just use pd.from_csv / from_excel to make it work. This video would be really short
Умничка!
where is the code kharcosde?
very confusing wording around the part non-dominated sorting. not very clear
Could you elaborate what you mean? Perhaps we can add the confusing part to the description
The non-dominated sorting algorithm/explanation kind of seems off.
If I'm not mistaken, your 1st example is not what you described as functions: when x1, x2=0, F1=0 and F2 =-8. Both F1 and F2 are ranging form -8 to 0....
What an amazing video! Saved me tons of time. Thank you :)
thank you its very helpfull until now...but it did not work with me..i got empty
encore merci j'avais bien galéré à l'installer sur windows docker va ma sauver la vie ... thank you again i had a hard time installing it on windows docker will save my life ;-))
Great video😊
Thank you for the great tutorial. In the code, you set "n_obj = 2", why there are two objectives? You only have one objective function "benchmarks.kursawe()"?
kursawe always returns two values. Therefore the number of objectives is 2 and it has 3 input params. See here: en.wikipedia.org/wiki/Test_functions_for_optimization
but im getting this output { "ok": true, "result": [] }, my result was empty, any allow settings me
you need to make your channel/group public, not private
Well explained!!!!