0:00:00.099,0:00:05.959 Hi! I'm Ian Witten from the beautiful University of Waikato here in New Zealand, and I want 0:00:05.959,0:00:10.560 to tell you about our new online course: Advanced Data Mining with Weka. 0:00:10.560,0:00:14.690 If you liked the other courses--Data Mining with Weka and More Data Mining with Weka--you'll 0:00:14.690,0:00:15.680 love this new course. 0:00:15.680,0:00:21.150 It's the same format, the same software, the same learning by doing, and the aim is the 0:00:21.150,0:00:27.150 same, too: to show you how to use powerful techniques of data mining on your own data. 0:00:27.150,0:00:31.029 One difference is that the lessons in this course are given by different people. 0:00:31.029,0:00:34.730 In fact, you'll get to meet pretty well the whole Weka team in this course. 0:00:34.730,0:00:36.969 This new course is advanced. 0:00:36.969,0:00:38.760 We're going to be looking at new kinds of data. 0:00:38.760,0:00:43.979 We're going to be looking at time series, for example, where the data evolves over time 0:00:43.979,0:00:46.100 and your job is to predict the future. 0:00:46.100,0:00:51.909 Or situations where the characteristic of the source changes slowly over time, 0:00:51.909,0:00:55.760 like it does in real life, and your job is to track those changes. 0:00:55.760,0:00:59.249 We'll look at different ways of working with big data. 0:00:59.249,0:01:04.680 We're going to introduce you to Weka's big sister, Moa, which is a stream-oriented data 0:01:04.680,0:01:09.950 mining system that never stores the data in main memory, so it can operate on effectively 0:01:09.950,0:01:17.000 infinite streams of data and has to use special algorithms to deal with these streams, 0:01:17.000,0:01:19.100 which we'll explain to you. 0:01:19.100,0:01:26.659 We'll also show you how to deploy Weka on a cluster computing environment using the Apache 0:01:26.659,0:01:30.820 Spark framework, and also the popular Hadoop framework. 0:01:30.820,0:01:34.969 We're going to show you how you can reach out to other data mining systems from Weka, 0:01:34.969,0:01:38.899 for example, the popular R data mining system. 0:01:38.899,0:01:44.359 You can get at all the algorithms in R, all the mining algorithms and all the 0:01:44.359,0:01:49.630 very powerful information display capabilities right there within your Weka interface. 0:01:49.630,0:01:55.060 We'll look at scripting Weka in Python, and you can write little Python scripts right 0:01:55.060,0:01:57.499 there in the Weka interface. 0:01:57.499,0:02:03.409 We'll show you how to set up the Python Weka wrapper, where you can access the Weka API 0:02:03.409,0:02:06.420 right from within your very own Python program. 0:02:06.420,0:02:09.710 By popular demand, we've included some applications. 0:02:09.710,0:02:15.350 We'll talk about the application of Weka to soil sample analysis, where machine learning 0:02:15.350,0:02:20.390 can effectively replace time consuming wet chemistry. 0:02:20.390,0:02:27.670 We'll look at signals from your brain, functional MRI signals, which treats the brain as a set 0:02:27.670,0:02:31.990 of voxels that extends over time, a kind of four-dimensional dataset of what's happening 0:02:31.990,0:02:35.530 in your brain and how to analyze this kind of data. 0:02:35.530,0:02:39.420 We'll look at a bioinformatics application: signal peptide prediction. 0:02:39.420,0:02:45.130 We'll look at some image processing stuff, some filters for getting features off images; 0:02:45.130,0:02:51.860 and we'll look at a Twitter application where you use Weka to do text mining on a Twitter 0:02:51.860,0:02:53.140 feed. 0:02:53.140,0:02:57.290 This course is aimed at teaching you the principles and practice of data mining. 0:02:57.290,0:02:59.980 We don't look at the technical details of particular algorithms. 0:02:59.980,0:03:04.120 In fact, you don't need any special mathematical background to do this course. 0:03:04.120,0:03:07.310 Indeed, you don't need any programming background. 0:03:07.310,0:03:10.730 We'll see some little Python programs, but you can pick that up along the way. 0:03:10.730,0:03:12.370 We'll show you how, don't worry. 0:03:12.370,0:03:16.320 Other things for this course ... You're going to need a computer, of course. 0:03:16.320,0:03:20.010 You're going to need an internet connection. 0:03:20.010,0:03:23.050 You're going to need a Google account, because again we're using the Google infrastructure 0:03:23.050,0:03:24.130 to deliver the course. 0:03:24.130,0:03:26.760 You will need a few hours a week each week. 0:03:26.760,0:03:29.490 You'll need a lot of motivation. 0:03:29.490,0:03:30.380 This is difficult stuff. 0:03:30.380,0:03:32.110 You're going to learn a lot. 0:03:32.110,0:03:33.700 The course lasts five weeks. 0:03:33.700,0:03:38.420 There are six lessons each week, and each lesson comprises a short video like this one, 0:03:38.420,0:03:43.540 followed by an activity where you get to practice what you've learned on a dataset that we provide. 0:03:43.540,0:03:44.790 And there are a couple of tests: 0:03:44.790,0:03:50.640 if you do well enough in those, you'll receive a statement of completion from the University 0:03:50.640,0:03:52.770 of Waikato, signed by me. 0:03:52.770,0:03:55.820 What else? Textbook? There is no textbook. 0:03:55.820,0:03:58.100 This stuff isn't in the books. 0:03:58.100,0:04:03.460 And I've recorded a new piece of music for you, some improvisations on a jazz theme by Dizzy Gillespie. 0:04:03.460,0:04:06.870 And the price of admission is zero. 0:04:06.870,0:04:08.830 This course is absolutely free. 0:04:08.830,0:04:10.870 So that's it. 0:04:10.870,0:04:15.370 Advanced Data Mining with Weka, coming soon to a computer near you. 0:04:15.370,0:04:16.970 Hope to see you there. 0:04:16.970,0:04:17.660 Bye for now!