One has to constantly remake themselves to adapt to the market and to the times. This activity serves those who practice it well economically and keeps them from becoming bored. It requires a constant state of personal “revolution”.
Here I describe my revolutionary plans for 2015; namely which technologies I intend to learn and which actions I will consider for the year.
Synthetic biology has the potential to revolutionize manufacturing, first the manufacturing of drugs and biofuels, then the manufacturing of more exotic materials. To do this we (researchers and engineers) will have to learn how to model biological circuits effectively (see my recent post “iBioSim: a CAD package for genetic circuits“). My intention this year is to strengthen my knowledge of synthetic biology’s computing side, leveraging my experience as an engineer and a bioinformatician in the process. Particularly I intend to combine this effort with my work to improve my statistical and machine learning knowledge, as both can be applied to the identification of genetic circuit parameters.
(Also see my post “synthetic biology: an emerging engineering discipline“) for a description of where the subject is at the moment.
Unstructured data is the frontier in data science. Currently such data (images, prose) is called “dark data” because it is difficult for a computer algorithm to automatically derive actionable information from it. However, new tools are becoming available for dealing with unstructured data. I therefore intend to learn how to deal with the kind of data found in prose: For example, extracting useful biological information from a bulk analysis of PubMed abstracts or beneficial stock tips from financial writing.
Internet of Things
I have no plans to deploy any Internet of Things objects this year, but am certainly interested in working with the data made available by such devices. If someone or some company wants to partner with me on this, I’m game.
Machine learning is becoming easier due to new tools and software. I want to get better at choosing among and applying machine learning techniques in the next year. I have already started this effort with two blog posts on clustered stock market analysis (“clustering stocks by price correlation, part 1” where I explored use of hierarchical clustering and “clustering stocks by price correlation, part 2” where I applied k-means clustering).
Particularly, I want to learn the intersection between Big Data and machine learning, e.g., how to combine basic Apache Spark with Spark’s MLlib toolkit.
I see opportunity to combine machine learning efforts with all the above-stated subjects, since my treatment of those subjects will be highly data driven.
Thinking Big: Warp Equations
I’m no physicist and certainly won’t become one in a year, but I want to start investigating the state of space-folding equations and related ideas. In particular I want to examine how close the ideas are to trial by commercial venture.