Once the initial motivation has worn off, it is hard to stay focused. In the end, who would blame you if you just stopped? The nice thing about the headstart into a new topic is the immediate reward in the form of new knowledge, understanding formerly complicated looking things, and being able to brag about starting a new thing on the next party. After a while, everything is back to normal. Nobody will ask anymore and you are left alone with your motivation.
When encountering the first obstacles, it is very likely that also your motivation will say goodbye and travel on to new projects and paths. If you want to become better than the average beginner, it is crucial that you work your way through this phase.
In my case, it was when I encountered problems in my machine learning project during the data preparation. Everything went well until then – I had identified a project that I really wanted to pursue, I had found enough data, I even had data available to personalize the project. Drunk from all these successes, my brain automatically jumped to the conclusion that this would be easy-going. Data cleaning and preparation is just a step on the ladder to my first ML model. Following tutorials that kind of matched my case, I assumed I would not run into any issues.
My data just didn’t fit. I realized that in most of the examples online, people already had numerical values as their data. I didn’t. I had directors, actors, genres. Even the year of release was not numerical, but a string value. In the tutorials online, it sounded quite easy to transform them. But in order to do so, you have to know into which values to transform them. Additionally, I knew that something was wrong with my data collection, because I was missing some values from my personal data. Dammit! What do you google when you don’t know your problem?
Keep calm and go on
After wasting two entire days on trying to figure out what cause my faulty data, I gave up. Not entirely – just for the time being. I couldn’t find it out, and as the husband was horribly busy that week, he couldn’t give advice either. I literally felt all my energy being drained from my body, and with it, all my motivation to go on. If I couldn’t solve this little issue, why should I go on at all? Here’s the reason: Because when focusing on my goal, it actually didn’t matter.
My goal was to go through all the steps of machine learning, up until I had a valid model that could be tested. With this definition in mind, it was not at all important to actually get a model that was perfect from the beginning. After letting the project rest for some days (yes, it took me several days to calm down), I knew that the best I could do was to note down all the open questions and go on with the next step. After all, I had a database now, with most of the data represented.
This shows how important it is to set your goal correctly, and to remind yourself at each step of your project that you should always focus on the goal. The real world is messy and there will be plenty of distractions. The mean thing is that many of those distractions will not look like it – instead, there will be a tiny voice in your head that shouts “But this is important, too!” – and it is correct! There are so many things that are important. However, if you have a goal, and encounter issues that will distract you from this goal, it doesn’t matter at all how important they look. Distraction is very subjective to what you want to achieve. If you set your goal right from the beginning, you will have an easier time identifying which ideas on your way to happiness are distracting, and which will contribute to your project. Don’t be afraid to say no – you can always note good thoughts down and come back to them later, when you achieved your goal.