Week 9: Finishing Touches and Checks

Rohan V -

Hi all, it’s Rohan!

Last week, I delved into some of the choices I was making regarding how to best represent my data and communicate results to a wide audience.

Now, I’ve completed the first draft of my paper; and I like how the results look so far. I mentioned last week that I was leaning towards focusing on three representative molecules of different length in my paper; this would show how the new change point detection algorithm is able to function well even across molecules with different characteristics. I’ve ended up following up on this goal, with a slight pivot, which I’ll explain below:

Although it’s useful to visualize how my change point detection (CPD) algorithm performs on a whole dataset, what’s arguably more important is how the algorithm works on a single graph of conductance against inter-electrode distance. Consequently, I decided to split my results section into two parts: one describing more ‘micro’ results, and the other ‘macro’ part describing how the algorithm works across a whole dataset. These two parts fit together quite nicely in that I’m able to delve into the specifics while avoiding redundancy (rather than explaining how my CPD algorithm works repeatedly, I can instead explain it one time, give a broad overview of its functionality, then show how it works on an entire dataset).

As a preview of what I’ll be discussing in my paper/presentation, one example of an application of my CPD algorithm to experimental output data is shown below:

Above, the image on the left shows the raw data, and the image on the right shows the data with the CPD algorithm applied.
As a refresher, the algorithm looks for points of significant structural change within conductance-distance data. The outputs are helpful for characterizing the conductance behavior of organic semiconductor molecules when they are placed in electronic devices.
Over the next few weeks, I’ll be finalizing my presentation, revising my paper, and creating a poster for display on presentation day. I’m excited to showcase more of these ‘traces,’ and show how my algorithm performs across broader ranges of data!

More Posts

Comments:

All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.

    adam_b
    Hi Rohan, your project is really fascinating! Recently some of my own work has involved assessing graphs so I'm wondering what software you used to make that graph?
    rohan_va
    Hi Adam, great question! I use matrix laboratory (MATLAB) to make these graphs. I have a set of x and y data, and then with the output from my change point detection function I can overlay the detected changepoints.

Leave a Reply

Your email address will not be published. Required fields are marked *