- 54
- 96 812
Enlightenment
United Kingdom
เข้าร่วมเมื่อ 14 ต.ค. 2014
This channel shares useful information on topics on life, and education. Please subscribe to get regular updates.
Architecting for success with Artificial Intelligence - Dr Steve Gustafson - CTO Noonum
#AutoML #MLOps #ArtificialIntelligence
Dr Steven Gustafson discusses what the difference is between success and failure of Artificial Intelligence projects in industry and presents a case study. He also introduces a framework for AutoML that he has invented and published.
Dr Steven Gustafson received his PhD in Computer Science and Artificial Intelligence, and shortly thereafter was awarded IEEE Intelligent System's "AI's 10 to Watch" for his work in algorithms that discover algorithms. For 10+ years at GE's corporate R&D center he was a leader in AI, successful technical lab manager, all while inventing and deploying state-of-the-art AI systems for almost every GE business, from GE Capital to NBC Universal and GE Aviation. He has over 50 publications, 13 patents, was a co-founder and Technical Editor in Chief of the Memetic Computing Journal. As the Chief Scientist at a Knowledge Platform software company, he invented and architected new AutoML and NLP microservices for industrial customers while growing an international data science and research team. With a passion for solving problems and innovation, Steven is excited to talk about what makes an AI project a success in industry.
Dr Steven Gustafson discusses what the difference is between success and failure of Artificial Intelligence projects in industry and presents a case study. He also introduces a framework for AutoML that he has invented and published.
Dr Steven Gustafson received his PhD in Computer Science and Artificial Intelligence, and shortly thereafter was awarded IEEE Intelligent System's "AI's 10 to Watch" for his work in algorithms that discover algorithms. For 10+ years at GE's corporate R&D center he was a leader in AI, successful technical lab manager, all while inventing and deploying state-of-the-art AI systems for almost every GE business, from GE Capital to NBC Universal and GE Aviation. He has over 50 publications, 13 patents, was a co-founder and Technical Editor in Chief of the Memetic Computing Journal. As the Chief Scientist at a Knowledge Platform software company, he invented and architected new AutoML and NLP microservices for industrial customers while growing an international data science and research team. With a passion for solving problems and innovation, Steven is excited to talk about what makes an AI project a success in industry.
มุมมอง: 251
วีดีโอ
R basics: File System - Variables - Vectors - Factors - Getting Help
มุมมอง 852 ปีที่แล้ว
#R #RProgramming #Vectors #Factors #Basics
Installing R, R Studio and R Packages
มุมมอง 1762 ปีที่แล้ว
R and R studio are widely used for statistics, data mining, optimisation and artificial intelligence.
Grammatical Evolution
มุมมอง 3K3 ปีที่แล้ว
This is a series of videos on Modern Optimisation methods. This lecture introduces Grammatical Evolution, an evolutionary automatic programming method to evolve computer programs in an arbitrary language. This lecture assumes familiarity with simulated evolution especially Genetic Algorithms. GP Bibliography: gpbib.cs.ucl.ac.uk/ A Field Guide to Genetic Programming By Poli, Langdon and McPhee F...
Particle Swarm Optimisation
มุมมอง 11K3 ปีที่แล้ว
#PSO #Swarm #ParticleSwarmOptimisation This is a series of videos on Modern Optimisation methods. This video introduces particle swarm optimisation. Predator Confusion Effect [1]: www.sciencedirect.com/science/article/pii/S0003347284802328 Many eyes hypothesis [2]: www.sciencedirect.com/science/article/pii/0003347295801499 Beni, G., Wang, J. (1993). "Swarm Intelligence in Cellular Robotic Syste...
Feature Selection With Genetic Algorithms - Code and Plots
มุมมอง 18K3 ปีที่แล้ว
#GeneticAlgorithms #FeatureSelection #Regression link to all the R code used: www.dropbox.com/sh/k8j696nfvjqp3h2/AAB2L3qgjZlf-_1beaWA_FC3a?dl=0 To extract the best linear model (alongside all its coefficients, and error analysis) use this code: www.dropbox.com/s/uxqr2dmioa30c53/getBestLmModel.R?dl=0 Linear Regression: th-cam.com/video/VtYqxjDhP7E/w-d-xo.html Research article (statistical signif...
Population Based Methods - Genetic Algorithms
มุมมอง 3.5K3 ปีที่แล้ว
#EvolutionaryAlgorithms #GeneticAlgorithms #Optimisation This is a series of lectures on Modern Optimisation Methods. “Exploration and exploitation in evolutionary algorithms: A survey” 2013. ACM Computing Surveys. dl.acm.org/citation.cfm?id=2480752
Single State Methods in Optimisation
มุมมอง 1.3K3 ปีที่แล้ว
#Optimisation #HillClimbing #SimulatedAnnealing This is a series of videos on modern optimisation, Genetic Algorithms and Evolutionary Algorithms.
Gradient Descent and Local Search
มุมมอง 1.6K3 ปีที่แล้ว
#EvolutionaryAlgorithms #GradientDescent #LocalSearch This is a series of short videos on Modern Optimisation methods such as Evolutionary Algorithms, Genetic Algorithms, and Genetic Programming.
Introduction to Modern Optimisation
มุมมอง 2.9K3 ปีที่แล้ว
#GeneticAlgorithms #EvolutionaryAlgorithms #Metaheuristics This is a series of short videos on Modern Optimisation methods. en.wikipedia.org/wiki/Mathematical_optimization Atari Game playing: link.springer.com/chapter/10.1007/978-3-319-55696-3_5
Linear Regression
มุมมอง 5164 ปีที่แล้ว
Regression: onlinecourses.science.psu.edu/stat501/node/250 Correlations - Pearson, Spearman, Kendal’s Tau: th-cam.com/video/IXJ1hXuo3HQ/w-d-xo.html ANOVA: onlinecourses.science.psu.edu/stat501/node/265 Based on the least squares criterion, which equation best summarizes the data? The sum of the squared prediction errors is 766.5 for the dashed line, while it is only 597.4 for the solid line. Th...
Correlations - Pearson, Spearman and Kendal's Tau
มุมมอง 6374 ปีที่แล้ว
Cancer mortality plots: onlinecourses.science.psu.edu/stat414/node/276 Statistical Hypotheses: th-cam.com/video/albAw5rMUkY/w-d-xo.html Parametric tests - Z-tests and t-tests: th-cam.com/video/2T7TAFl3zvA/w-d-xo.html Non-parametric tests : th-cam.com/video/hWHZz7yGbno/w-d-xo.html Anscombe's quartet: en.wikipedia.org/wiki/Correlation_and_dependence#Correlation_and_linearity Spurious Correlations...
Non parametric hypothesis testing
มุมมอง 7114 ปีที่แล้ว
Statistical Hypothesis: th-cam.com/video/albAw5rMUkY/w-d-xo.html Central Limit Theorem: th-cam.com/video/5O3FOVAlNfA/w-d-xo.html Descriptive Statistics: th-cam.com/video/RU0m22JJ1vQ/w-d-xo.html sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs704_nonparametric/BS704_Nonparametric2.html Complete Exercise sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs704_nonparametric/BS704_Nonparametric4.html Reject H0 if : ...
Hypothesis testing: Z-test and t-test
มุมมอง 7844 ปีที่แล้ว
Statistical Hypothesis: th-cam.com/video/albAw5rMUkY/w-d-xo.html Central Limit Theorem: th-cam.com/video/5O3FOVAlNfA/w-d-xo.html Descriptive Statistics: th-cam.com/video/RU0m22JJ1vQ/w-d-xo.html #Z-test (R code): z.test = function(a, mu, var){ zeta = (mean(a) - mu) / (sqrt(var / length(a))) return(zeta) } #Sample call a = c(65, 78, 88, 55, 48, 95, 66, 57, 79, 81) z.test(a, 75, 18) #Sample output...
Statistical Hypothesis
มุมมอง 6924 ปีที่แล้ว
Why Say Fail to Reject in a Hypothesis Test? www.thoughtco.com/fail-to-reject-in-a-hypothesis-test-3126424 Introduction: This is a series of videos that produces a university course on Applied Statistics. The link to video series is: th-cam.com/play/PLU98PJIZ0JfJHi1dnjpcHz6_Fy5zUOZg_.html The topics covered are: Aggregating information; measures of central tendency. Spread of information; data ...
Sampling Distribution and Central Limit Theorem
มุมมอง 5514 ปีที่แล้ว
Sampling Distribution and Central Limit Theorem
Descriptive Statistics - Measuring central tendency and dispersion; histograms, box plots and more
มุมมอง 2.1K4 ปีที่แล้ว
Descriptive Statistics - Measuring central tendency and dispersion; histograms, box plots and more
Automatic Programming with Genetic Programming
มุมมอง 1.2K4 ปีที่แล้ว
Automatic Programming with Genetic Programming
Powerful Paragraphs 12: Continuity - Link Key Terms
มุมมอง 2085 ปีที่แล้ว
Powerful Paragraphs 12: Continuity - Link Key Terms
Ejaculated
❤
EXCELLENT! Would it be ok to use logistic regression for the fitness function in the case of classification problems?
Best Video in Modern Optimization ❤.
🎉🎉🎉🎉 Graet video.. R has awesome package Gramevol for grammatic evolution
How does it work with laser pm 2.5 particle sensors ? What’s the objective
So so sad. In logo ko hata k pesudo tabdeeli sarkar ko laya gaya where Khan made sure to ruin everything down the drain through Buzdar
thanks for your efforts.this is a hybird feature selection via GA,is that true?
Really good, thanks
Reall impressive
Hi, have you ever used other regression equations such as PLS regression or ANN regression as the fitness function?
👍👍👍☺
MashaaAllah
MashaaAllah
Good 👍
MashaaAllah
Ok. You have use WHOEVER before a verb: Whoever finds my cellphone will receive a reward. You have use WHOMEVER before a noun or pronoun: I would like to go out with whomever my friends suggest. The most simple rule in order to learn this grammar point.
Thanks so much for this :)
Can you please share the data used in this video in CSV format?
Kelvin is a Ratio Variable...
Great
This guy is Corrupt and has lot of Money laundered from NGOs
Many people want to know how to get the best model at the end. See the code below: : www.dropbox.com/s/uxqr2dmioa30c53/getBestLmModel.R?dl=0
thankkk you!! this series is tooo much help
What's your country name 🤞🏻🤞🏻
can i know why the AIC value get even higher over generation? i thought the better regression model have lower AIC. great explanation anyhow!
Thanks for this clarification, but I still don't understand from where did you get the numbers 60 - 14 - 64 - 125 - 65 - 125 - 128 - 233? Could you clarify this for me please?
GE uses a GA to generate chromosomes, that is, bunch of numbers. In the first generation, the numbers are randomly generated up to a given length. This is just an example sequence of numbers that could have been generated by GE. Makes sense? If you are not clear on Genetic Algorithms, then watch this: th-cam.com/video/CaYXJtSbcXs/w-d-xo.html and its highly popular implementation with code th-cam.com/video/s5uYrvL_Hkw/w-d-xo.html
Very nice presentation indeed
Amazing explanation. Is there any command to show the summary of the fittest linear regression? (like its regressors coeffiecient, p-value, etc.)
Happy to help. You can get the fittest model using the code here: www.dropbox.com/s/uxqr2dmioa30c53/getBestLmModel.R?dl=0
Thanks Dr Atif.
can I get codes in python?
Hello, Thank you... you have a nice video presentation. I have a question, is it possible to identify the boundaries in PSO?
Hi Denis. It depends what you mean by "boundaries"? If you mean boundaries of the search space, then the user (problem provider) defines them. If it something else, please explain.
@@enlightenment609 Yes, I mean the boundaries of the search space, in the equation I cannot determine what is the boundaries of the search space. If none, Can you please help me identify where to put in the equation or in the algorithm? Thank you in advance.
@@dennisbarrios7562 those boundaries are dettermined by the user. For example, suppose you are optimising some pressure values. Then you would know how far up or down the pressure can be turned. So the user knowledge determines those boundaries. Similarly, say you want to set working hours then you would know what is the minimum and maximum working hours. So you can think further examples like that.
Thank you for Helping me in tomorrow's exam ❤
happy to be of some help.
Thanks!! a lot!!
Thank you very much.. Please excuse me. I would like you if you would explain to me the possibility of modifying the chromosome.. I am working on a master's thesis in the field of predicting missing values. Can you help me with a code that predicts them? Or you can give me a reference that I can benefit from.. and I repeat my thanks to you
Your comment does not clarify why and how do you want to modify the chromosome just to predict missing values (from what?). ) If you want to build a classifier that predicts if a dataset/record has missing values then GA must only select features for such a classifier. In this case the same binary chromosome applies.
Whoever made this video is a legend.
So kind but this is only a humble attempt.
You are honestly really underrated. Please do post more videos and keep growing . I've check out previous videos and they were so clear cut and to the point. Thank you!
Thank you @aditya. I have a few in the pipeline but just need to find time to prepare them. Will do it as soon as I can.
Great illustration. thanks
Glad you liked it!
The allegation fostered investigation. Students discussion of the topics, proved their engagement with the lecture.
The average score of the collaborative learners at the end of the semester was higher than that of the solo learners. The patient's vital signs did not improve, despite medication.
Results improved Performance evaluation happened Account audited
14.52 ( The value will be the same as the value of the slope will become 0 when x is already at maximum or minimum)
I think in the car price question it depends on the purpose of the study
bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
Great sir ❤️👍
Love it! Thank you...can you do a video on how to train neural network using PSO?
Its actually not that difficult. PSO/GA etc can be used to tune the hyper parameters of a Neural Network; the training itself will be internal to the NN. To do this, you simply plug in the NN into the fitness function of a PSO and then initialise it with the PSO generated values of the hyper parameters.
A. Either Operator Equalisation (an example of bloat control method) or Nonparametric Bloat Control (another example of bloat control method) can be used in Genetic Programming. In fact, the two bloat control methods are well known.
B. is better: B. There are many approaches to studying the dynamics of science. Some approaches use statistical methods; however, other approaches, termed computational history of science, use computational methods.
Internet-based work allows many opportunities to grow earnings. Internet is proliferating, and internet-based startups and remote workstations allow one to work from anywhere; the kind of work available on the internet includes jobs that require high skills. A high household burden often disallows women from working. Also, in some eastern societies, highly educated women who are willing to stay at home after marriage are highly sought after as brides. Since marriage is near critical for women to function well in such societies, women often stay unemployed after marriage.
A good attempt. You can improve it further by putting the second paragraph first to explain what the problem is. Also, I would start the (then) second paragraph with "However", just to link the first para (with the problem statement) with the second.
@@enlightenment609 A high household burden often disallows women from working. Also, in some eastern societies, highly educated women who are willing to stay at home after marriage are highly sought after as brides. Since marriage is near critical for women to function well in such societies, women often stay unemployed after marriage. However, Internet-based work allows many opportunities to grow earnings. Internet is proliferating, and internet-based startups and remote workstations allow one to work from anywhere; the kind of work available on the internet includes jobs that require high skills.
1. I saw a picture whilst walking past the shop. 2. Resolving confusions wastes the reader's time. The action of wasting the reader's time must be avoided. 3. Individual neurons respond to stimuli only in a restricted region of the visual field; known as the Receptive Field. 4. Weekly exercise of less than 1 hour did not improve health, 3 hours improved a little, and 5 hours improved significantly. 5. The lectures were finished and the assignment was submitted. 6. The lectures were finished and the assignments were submitted before the respective deadlines.