Understand how machine learning is being used in a range of practical situations today
Machine learning is an area of computer science that aims to help machines learn by example, to adapt to new circumstances and to detect and extrapolate patterns. For example, a machine may be shown pictures of apples and oranges and initially told what kind of fruit that photo contains. Over time, through machine learning software, the computer is able to reliably determine what the photo represents without being told and without human input. The machine learning software operates much like human learning; the software learns to determine what features are unique to the orange and the features unique to the apple, e.g. colour, shape, texture, size etc., to reliably predict the fruit type.
Machine learning developed from the study of pattern recognition and it involves the creation of algorithms or software code that can learn from, and make predictions on, data. It also has ties to statistics and mathematical optimisation and is used in situations where designing and programming explicit rule-based instructions is not feasible. Machine learning techniques are increasingly being used to handle so-called “big data” where the amount of information being collected is so large and varied that traditional data processing techniques are inadequate.
Machine learning often uses something called artificial neural networks which are interconnected families of algorithms or models, inspired by the human brain, where particular algorithms get turned on depending on the data fed in and then exchange messages with other algorithms in the network. The connections or pathways within the neural network have numeric weightings which are tuned, based on experience with different inputs, so they are capable of learning. Using neural networks in machine learning allows complex tasks to be broken up and solved in parallel by different parts of the network, rather than sequentially, and the improvements in computer power have made artificial neural networks feasible.
Machine learning is being used in a range of practical situations today. For example, Google introduced machine learning in its Google Now speech recognition system (which is Google’s version of Apple’s voice-driven virtual assistant called Siri). By using machine learning techniques, Google was able to cut error rates by 25%. Facebook uses machine learning for facial recognition to identify and tag people in photos that its users upload. Amazon uses machine learning to set prices on its 20 million products by making price comparisons against competitors, or if the product is not identical to one stocked by a competitor, it looks for similarities and then prices against similar products that competitors stock. Machine learning is also being used to create driverless cars and drones by allowing the vehicles to identify objects and learn from their surroundings.
Another practical application of machine learning and artificial neural networks is in cancer diagnosis,Systems that use these techniques have been used to detect or predict a range of cancers by looking at gene information in large groups of patients and comparing the information with a single patient to reliably predict whether the patient is likely to have cancer. Some technologists looking at longer term technology trends are already forecasting that machine learning techniques will be able to find and test new drugs at a much faster rate than is currently the case and that machine learning will ultimately allow machines to be able to program themselves, display forms of artificial intelligence and replace many of the blue and white collar jobs of today through robotics and automation.
This article is prepared for information purposes only. It does not purport to contain all matters relevant to any particular investment or financial instrument, or to a general investment approach.
This article contains general advice only. It does not take into account your investment objectives, financial situation or particular needs. You should consider whether this advice is suitable for you and your personal circumstances.
JBWere Ltd (‘JBWere’) and its related entities distributing this article and each of their respective directors, officers and agents (‘JBWere Group’) believe that the information contained in this article is correct and that any estimates, opinions, conclusions or recommendations contained in this article are reasonably held or made as at the time of compilation. However, no warranty is made as to the accuracy or reliability of any estimates, opinions, conclusions, recommendations (which may change without notice) or other information contained in this article and, to the maximum extent permitted by law, the JBWere Group disclaims all liability and responsibility for any direct or indirect loss or damage which may be suffered by any recipient through relying on anything contained in or omitted from this article. JBWere Group may provide investment recommendations, market commentary or trading strategies to clients reflecting opinions that are contrary to the opinions expressed in this article, or that are inconsistent with the recommendations or views expressed in this article.
This article may contain information based on JBWere’s understanding of taxation and other laws. JBWere does not hold itself out as providing professional taxation advice. You should consult with your professional taxation advisor before acting on the information or data contained in this article or contact a financial advisor if you require further assistance