"Are you happy with what you have done and is doing in your work?" is the key. Reading "The Passion Paradox" let me realize if we are art-crafts, the point is how we demonstrate what we want to show instead of the price we are valued. While on the other side, still there will be some people disagree with it. Sometimes, only when some KOL (Key Opinion Leader), like let's say FAANG companies, recognized us with the values, then most people will agree with the price. There are definite some people think we are over-valued and people thinking we are under-valued. However, I feel like we are all art-crafts in the job market. Myself also have this doubt many of the time. This kind of imposter syndrome is really frequently seen. One of my friend told me that he's worrying that he doesn't worth how much he's paid right now. While I'm not quite sure if that's the truth or there are other reasons, would like to know if anyone have answers toward this question! And thus, it becomes a rule of thumb to set alpha as 0.05 and beta as 0.2. I'm assuming if you set alpha to 0.01 and beta to 0.1, the sample size needed will increase in a exponential way. The first thing I think of is that it's probably trying to balance between precision and recall meanwhile make sure the needed sample size go up significantly. He asked, "Then do you know why we always set alpha as 0.05 and beta as 0.2 to have 80% of statistical power?" I pause for a second because I never really think about that. However, when I discussed with my colleague, Tino Cheng, about this topic. The variance of target variable: so that you'll have an idea if it's a true effect or noise, seasonality, etc MDE: The minimum detectable effect you want to set. beta: Related to statistical power, which can be similar to recall metric. alpha: Decide the p-value threshold you want to control your true positive and so is precision. And the answer I can think of is that there are: I read a question asking what factors will affect needed sample size in AB Testing. We may not be paying the most, but some decent growth for sure. If you are interested in working at Houzz, feel free to let me know. Therefore, I look forward to the upcoming challenges and the growth along. “It’s always uncomfortable because you are going upward.” One of my seniors once told me. While it’s true that I’ve been stressful for not doing good enough many of the time, I think it’s also a process toward more responsibility and better capability. I really appreciate the opportunity as it leads me to find my passion for experimentation and causal inference in data science and great teams of both Marketplace and Taiwan team. Despite that we haven’t done a survey about satisfaction, I’ve got many warm messages from the team. And I’m glad that we’ve made some progress in the past few weeks. Therefore, earlier this year, I volunteered and fortunately got elected as one of the two Taiwan Labor Representatives to lead the communication with leaders. However, I think instead of complaining or leaving, it’s better to take initiatives and foster communications with leadership. Though this is super challenging to manage projects and communications with different stakeholders, I seriously think I learn and grow a lot in the process.Īnd as Taiwan is the new site of Houzz, we built many things from scratch and there are many things we don’t like about. And as the time passed, I take charge of more and more areas. To work on product development using analysis and AB testings, I tried my best to get ramped up as soon as possible and read a lot of references to know more about the field. It’s definitely the longest and the year I learned the most. Today marks my One-Year Anniversary at Houzz.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |