Poster Presentation

And so my final task, the dreaded Poster Presentation, journey’s end.

Was it a dread, NO;  I really enjoyed my few minutes of fame, presenting my research.  Even though my experiment had failed, I’d research everything, the history of machine learning, as well as potential futures.  I’d also looked at how I could take my experiment forward, making it proactive rather than reactive, so when that particular question came up, I was ready.

I had confidence not just it what it should’ve done, but the underlying reasons why it had failed, and how I could fix it in time for BCUR20.

Oh yes, BCUR20, I must fix it by then!

Alas, farewell, stay safe, and here’s to a very different future!

Crash, bang and wallop!

And just when you think you’re actually enjoying the process, things go horribly wrong!

I got to Step 3 of my methodology, after a long and complicated process of creating the final development environment, and wrangling my CSV files. All I needed to do was revise the Template Code I had researched, to fit my data sets, and crash!

The first problem was simply a case of deprecated code. A search and evaluation of its purpose have led me quickly to an updated solution.

Next, it simply failed to progress because it refused to identify the headers of my CSV! How could it not? A CSV is a simple text file, headers on the first row.

It took me a couple of days to understand that it was MY code at fault, but the underlying code, I felt like crying. Do I continue to fix this underlying code, wondering how deep down the rabbit hole would I need to go. I decided it was time for some professional academic advice.

After some hasty meetings with my Progress Tutor, Supervisor, and the Course Leader, admitting my problem; it was agreed that the experiment had failed; instead, I should focus on why? What could be done differently, and how might it have produced results.

Time for some rapid updates!

The COVID Affect

It has been a hard few weeks, since lockdown started on Monday, March 23.  I thought I’d be OK, time to just focus on what I needed to complete the course, how wrong was I.

The lockdown hit me much harder than I expected. To me, it was like falling into a whirlpool. No matter how hard I tried to swim to the edge, so I could climb out and get on with what I needed to do, the more I was dragged to the centre and down, unable to do anything. I now attribute these feelings, and the extreme writer’s block, to my loss of structure, and human contact. It is surprising how lockdown has impacted so many people in similar ways.

I took until today to accept that I needed help to complete my work, so I rewrote my timetable to be more realistic to my needs, and reconnected with my support network; something I should’ve done sooner. Fortunately, the B&FC accepted that I had been struggling, and agreed a PMC (Personal Mitigating Circumstances) giving me more time to complete two of my modules. I knew that the two upcoming exams MUST be done on the day, no matter what, so I spent more time preparing for these.

I feel happier and more focused now, time to march on.

Formative Submission Day

Well, it’s has done, submitted my Dissertation for a final review by my Supervisor, Michael Thornhill. After this, I’m on my own, cripes!

I am nervous because as usual, I’ve taken a different route to completion. I’ve built my Dissertation around the research I’ve done and the references I’ve collected, rather than just writing and then finding the References to support the arguments, which for me would take forever; the time I don’t have.

Looking back through the Dissertation now, I literally have pages and pages of references with in-text citations, all with comments. Hence, it’s just a case of expanding on those comments and linking it all together.

Linking it all together is where the Academic Phrasebank by Morley (2018) will come to the rescue.

Note to self: must progress with the experiment, not too worried as the complex development environment is already built, might need to upgrade a few libraries at coding time. Hence, as I said, not too worried.

References: –

Morley, J. (2018) Academic Phrasebank. 2018 Edn., University of Manchester. 2018 Edn. Manchester: University of Manchester. doi: 10.1093/ml/gcm073.

Dissertation Framework Complete

By merging the Computing Dissertation – House Style Template [19-20] with the format of Keith Whiteheads (Tutor) excellent Thesis, which he kindly allowed us to see, I have today completed, and I’m happy with, the structure of my Dissertation.

The structure did take some work to amend, because the House Style was for all streams of computing, and as such, not all was headings were relevant to my particular experiment.

TIP: When working on my Dissertation using Microsoft Word, I found that switching ON the SHOW/HIDE option on the HOME Ribbon….

….has helped me tremendously while formatting. This option shows all the hidden Word codes that structure the document, enabling me to fine tune the layout, and switch from Roman Numeral page numbers, to Page X of X in the main body.

BCUR20 Abstract Accepted

I’m in!

After a number of revisions, my submission to BCUR20 has finally been accepted.

Now I’ve another milestone to prepare for, but the additional support from Chris Willitts will help tremendously.

I plan to present an animated Powerpoint, with a voice-over; answering questions at the end.

How do I feel?  Suffering from that classic sensation of Impostor Syndrome; I keep reminding myself that I must be good enough because the panel decided my research was of interest.  My greatest fear is failing to deliver the Q & A at the end, I guess lots of practice with metaphorical naked people is on the cards.

Updated Dissertation Framework

Today I discovered that the Computing School had released the promised Computing Dissertation – House Style Template [19-20] onto Moodle.

I’ve downloaded this approved Framework, and intend to update my original Framework to bring it into line with this new format.

I’ve taken a quick look and found it a little bland, so intend to tweak it a little with a better cover page, style and colour.  Other than these minor tweaks, I do prefer this new format.

To be or not to be?

As I stated in my previous post of 12th December 2019, I can be distracted when doing research, often spinning off on a related tangent.  Such was the case on Sunday 5th January while downloading the software I required to set up my server yesterday.

I acquired the software from the Microsoft Educational Website, access provided by B&FC;  while looking for the software I needed, I noted that Microsoft has a stand-alone dedicated Machine Learning Server 9.4.7 for Windows, that offered an alternative route to the development of my experiment.

It was tempting, however, the learning curve was too steep, given the time I have remaining, so the idea of switching was shelved.  Its existence was noted for further investigation;  after my dissertation, I hasten to add.