Problems that ML is suited to solve
1. Problems that ML is suited to solve
Machine learning lets computer systems continuously adjust and enhance themselves as they accrue more experiences. For this reason, when more data is put into them, the results are more accurate. With this in mind, ML is suited to solve four common business problems. The first is replacing or simplifying rule based systems. Let's use Google Search as an example. Suppose you want to search for the Giants, a US sports team. If you type in Giants, should the search results show you the San Francisco Giants or the New York Giants? One's a baseball team based in California and the other is an American football team based in New York. In years gone by, the search engine used hand coded rules to decide which sports team to show user. If the query is Giants and the user is in the Bay Area, show them results about San Francisco Giants. If the user is in the New York area, show them results about NY Giants. If the user is anywhere else, show them results about tall people. This was for just one query. If you multiply this process by millions of different queries and users each day, you can probably imagine how complex the whole code base became. This is a perfect problem for ML to solve. If all the data that's available shows which search results users clicked on per query, a machine learning model can be trained to predict the rank for search results. A second business problem ML can help solve relates to automating processes. ML is designed to make predictions and repeated decisions at scale. Let's explore another example, this time from a property developer headquartered in Thailand called Ananda Development. For every sale, both an Ananda Development inspector and the buyer have to conduct a detailed check of the property. This was a manual, time-consuming process that was prone to much human error. Inspectors would visually check hundreds of items a day for problems, list any issues on paper and then photograph the findings. Multiplied across several projects, this workload adds up. Ananda Development decided to create a mobile application to make this process more efficient. Inspectors would verbally describe defects and critical issues to the application that ran on their smartphones. The application would then track and document the inspection results. In planning the application, the business realized it would need to recognize and convert to text, Thai language, speech and a version of English spoken by many Thai people. They decided to automate this process using Google's speech-to-text API. Furthermore, Ananda Development wanted to establish a pathway to use machine learning to complete condominium inspections by using remotely piloted drones. They decided to automate that process by using the Cloud Vision API to capture images of defects and automatically classify information about each one. Within three months of implementation, Ananda Development had saved around 130 hours of inspection time and over 100,000 US dollars in manpower costs. The inspection process is now more efficient and accurate. And as another benefit, buyers also receive copies of electronic inspection reports and updated status notes as defects are repaired. So far, you heard about ML problems that use structured data to make predictions at scale. A third type of business problem that ML can help solve is understanding unstructured data like images, videos, and audio. This example comes from Ocado, one of the world's largest online only grocery supermarkets. Previously, when Ocado received emails, they would all go to a central mailbox for sorting and forwarding by a human. This process was time-consuming and led to a poor customer experience. To improve and scale this process, Ocado used ML's ability to process natural language to identify the customer's sentiment and the topic of each message, so that they could route it immediately to the relevant department. This eliminated multiple rounds of reading and triaging, and ultimately improved customer satisfaction, and retention. And finally, there's personalization. Many businesses use ML to personalize user experiences and YouTube is a great example of personalization in action. When you watch a video on YouTube, you've probably noticed there's a list of recommended videos that are up next. When your video finishes, a new video will play and YouTube wants it to be interesting and useful for you. By using ML to provide personalized recommendations, YouTube can deliver a better customer experience. Many businesses use this same approach to surface product recommendations on their websites that are personalized to individual users. Other businesses use personalization to surface new content like music recommendations or films to stream. It's important to remember that ML models aren't standalone solutions and that solving complex business challenges requires combinations of models. There are of course, many more applications of machine learning for businesses and you can learn even more about them in our machine learning courses.2. Let's practice!
Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.