WEBVTT 00:00.080 --> 00:01.040 Welcome back. 00:01.080 --> 00:08.120 In general, machine learning algorithms are used in a wide range of applications like fraud detection, 00:08.120 --> 00:15.080 computer vision, autonomous vehicle predictive analytics where it's not computationally feasible to 00:15.120 --> 00:21.760 develop conventional algorithms that meet the requirements of real time and predictive nature of work 00:21.760 --> 00:23.560 in the field of machine learning. 00:23.600 --> 00:31.960 There are multiple algorithms that help us reach descriptive, predictive, and prescriptive result 00:32.000 --> 00:34.920 sets based on parameters defined. 00:34.960 --> 00:38.440 Descriptive explaining with the help of data. 00:38.480 --> 00:41.600 Predictive predicting with the help of data. 00:41.640 --> 00:45.320 Prescriptive suggesting with the help of data. 00:45.320 --> 00:52.360 We understand the functions of machine learning from the information and in the previous slides. 00:52.360 --> 01:00.000 But the art of making those functions work is in the way in which the algorithms are designed and used 01:00.000 --> 01:02.480 to execute those functions. 01:02.680 --> 01:08.060 Let's start with the first type, which is the supervised Learning. 01:08.220 --> 01:15.580 Supervised learning is a type of machine learning where a model learns from the labeled data, meaning 01:15.580 --> 01:18.860 every input has a corresponding correct output. 01:18.900 --> 01:26.140 The model makes predictions and compares them with the true outputs, adjusting itself to reduce errors 01:26.140 --> 01:28.420 and improve accuracy over time. 01:28.580 --> 01:33.420 The goal is to make accurate predictions on new, unseen data. 01:33.540 --> 01:41.380 For example, a model trained on images of handwritten digits can recognize new digits in digits it 01:41.380 --> 01:44.020 has never seen before. 01:44.220 --> 01:48.540 The second type is semi supervised learning. 01:48.700 --> 01:55.900 Semi-supervised learning system is like a supervised learning system, but it uses both labeled and 01:55.900 --> 01:59.300 unlabeled data in addition to supervised data. 01:59.460 --> 02:06.580 The term labeled data refers to information that has a meaningful tag that allows the algorithm to understand 02:06.580 --> 02:15.200 the data, whereas unlabeled data doesn't have such a tag, which means that ML algorithms can be told 02:15.200 --> 02:18.080 to label data that has not been labeled. 02:18.120 --> 02:22.480 Another type is the unsupervised learning. 02:22.760 --> 02:31.400 Unsupervised learning is a type of machine learning that analyzes and models data without labeled responses 02:31.400 --> 02:33.440 or predefined categories. 02:33.560 --> 02:40.320 Unlike supervised learning, where the algorithms and the algorithm learns from input output pairs, 02:40.440 --> 02:49.520 unsupervised learning algorithms work slow solely with input data and aim to discover hidden patterns, 02:49.520 --> 02:57.640 structures, or relationships within the data set independently, without any human intervention or 02:57.640 --> 03:01.120 prior knowledge of that data's meaning. 03:01.240 --> 03:04.560 Another type is reinforcement learning. 03:04.680 --> 03:13.040 Reinforcement learning RL is a branch of machine learning that focuses on how agents can learn to make 03:13.040 --> 03:18.360 decisions through trial and error to maximize cumulative rewards. 03:18.640 --> 03:27.140 RL allows machines to learn by interacting with an environment and receiving feedback based on their 03:27.140 --> 03:27.860 actions. 03:28.060 --> 03:32.300 This feedback comes into the form of rewards or penalties. 03:32.340 --> 03:32.940 Don't worry. 03:32.980 --> 03:36.220 We're going to learn about those in the next videos. 03:36.500 --> 03:44.340 And as a quick recap, we use the supervised learning for classification and regression, classification, 03:44.340 --> 03:51.700 fraud detection, email spam detection, diagnostics, image classification and in regression, risk 03:51.700 --> 03:54.260 assessment and score predictions. 03:54.300 --> 04:02.420 The unsupervised learning used for reduction, text mining, data visualization, face detection and 04:02.420 --> 04:03.580 voice detection. 04:03.740 --> 04:04.620 Regression. 04:05.060 --> 04:08.020 City planning and targeted marketing. 04:08.660 --> 04:10.260 Reinforcement learning. 04:10.540 --> 04:11.700 Finances. 04:11.740 --> 04:17.020 Manufacturing, stock management, autonomous cars, and others. 04:17.020 --> 04:23.420 In the next video, we're gonna deep dive into those types of ML algorithms, and we're going to learn 04:23.420 --> 04:28.700 about their nature and how it how they work.