Plan with me for final year of my PhD - Setting up my academic calendar

TL;DR

Use a “year at a glance” page to map conference deadlines, teaching/university weeks, business dates, and personal events across the whole year, then add monthly detail for execution.

Briefing Cornell Notes

Briefing

Final-year PhD planning centers on building a single, usable academic calendar that ties together thesis milestones, conference deadlines, teaching obligations, personal commitments, and the specific research work still left to finish. The core move is using a “year at a glance” page (inspired by bullet journaling) for high-level visibility across the next year, then breaking that into monthly views and—when helpful—mirroring key events in Google Calendar. That structure is meant to prevent the common problem of losing the “shape” of upcoming months when everything sits in a standard calendar grid.

The research schedule is anchored in a recommender-systems thesis focused on training recommendations for marathon runners using data from smart watches via the Strava app. Three major tasks remain: (1) revisit the raw dataset and extract additional features to better understand which training features correlate with marathon performance; (2) build a new prediction model using different methods to avoid being boxed into a single approach; and (3) integrate everything into a full recommender system, including a user interface component and user evaluation to strengthen the thesis. A smaller, less desirable add-on—returning to training descriptions work—is slated as a late-stage, non-publishable thesis enhancement.

Conference timing drives when each research block can realistically happen. The calendar begins with September, when an early presentation requires slide preparation within a week, followed by a 21-day holiday that includes a working week for the first in-person conference. After classes resume on the 12th, the workload shifts to multiple slide-deck deadlines for two presentations, then back to lecturing once materials arrive. In October, the first major academic target is the Intelligent User Interfaces conference (IUI): a discussion with the supervisor is planned first, then a full paper deadline of October 14. Because the data work for feature engineering and marathon prediction has not been completed yet, those tasks are pushed to after the IUI deadline, while the IUI slot becomes the focus for the full recommender system and UI validation.

Beyond that, the year is mapped around submissions and recovery time. A typical pattern is taking a week off after each submission; a week near Halloween is planned because other responsibilities ease. November is mostly reserved for a sister’s hen party, while December concentrates on drama school performances, finishing classes, a research studies panel meeting, and major personal events including a sister’s wedding, the birthday, and Christmas. A January submission is expected, after which thesis writing begins immediately with a goal of finishing by June. The writing order is designed to minimize rework: early chapters will reuse conference and journal paper work (data preparation, methodology, and results), followed by a full literature review around March, then the introduction, discussion, conclusions, and final revisions.

Teaching and drama school commitments continue throughout, with additional travel-linked work only if conference acceptances arrive—such as a Washington DC conference in February, Sydney in late March for IUI, and possible trips around Easter and early June. The plan ultimately balances research deliverables with realistic calendar constraints, so thesis writing can start on time and leave room for breaks after major deadlines.

Cornell Notes

The plan for the final PhD year builds a year-at-a-glance calendar that connects thesis work, conference deadlines, teaching duties, and personal commitments. Research priorities remain: extract and analyze new features from the raw Strava smart-watch dataset, train a new marathon-performance prediction model using multiple methods, and integrate these into a full recommender system with UI validation and user evaluation. A smaller “training descriptions” task is deferred to the end as a thesis add-on. Conference timing reshapes the order of work—feature engineering and prediction are pushed until after the October 14 IUI deadline, while the recommender-system and UI validation become the focus for that submission. Thesis writing starts in January and is targeted to finish by June, with early chapters based on already-written conference/journal material and a literature review scheduled for March.

How does the calendar system work, and why isn’t a single Google Calendar enough?

The system uses two main planning layers. First is a “year at a glance” page (built using bullet-journaling principles) that lists key dates across the next year—conference dates and submission deadlines, journal issue deadlines, teaching/university weeks (including breaks and class start dates), business dates for the classes and staff availability, and personal events like holidays, birthdays, and anniversaries. Second is a monthly spread that places those key dates into a more detailed month view; Google Calendar is used selectively for certain events. The stated goal is to maintain a clear sense of upcoming months rather than losing that overview inside a standard calendar grid.

What are the remaining research tasks for the recommender-systems thesis, and how do they fit together?

Three core tasks remain before wrapping research: (1) return to the raw dataset and extract additional features, including feature explanation and exploration to identify training features that correlate with marathon performance; (2) build a new prediction model using different methods tried previously, so the thesis can justify the final approach; and (3) bring everything together into a full recommender system from end to end, including a user interface component and user evaluation to strengthen the work. A small extra task—revisiting training descriptions work—is scheduled at the end as a thesis-only add-on rather than something expected to be publishable.

Why does the IUI conference deadline change the order of thesis research work?

The October 14 IUI full-paper deadline is treated as a forcing function. Because the data work for feature engineering and marathon prediction hasn’t been completed, those tasks are pushed to after the IUI deadline. Instead, the IUI submission becomes the target window for the full recommender system and the UI validation/evaluation work, while later months are reserved for the deeper feature and prediction modeling.

What does the September-to-October schedule look like in practice?

Early September includes an in-person presentation that requires slides within a week. A 21-day holiday follows, with a working week inside it for the first in-person conference. Classes resume on the 12th, and the early absence triggers a need to email lecturers about coverage/absence. During the holiday period, two presentations require additional slide preparation. Lecturing work resumes later in the month once materials arrive. In October, the IUI submission planning begins with a supervisor discussion, then shifts into paper preparation leading to the October 14 deadline.

How is thesis writing planned after submissions, and what order is prioritized?

A January submission is expected, and thesis writing begins in January with a target completion in June (with possible breaks depending on conference timing). The writing sequence is designed to reuse existing conference/journal outputs: early chapters focus on data set readiness, then methodology and results for each sub-project. Those chapters are expected to be easiest because corresponding papers already exist. A full literature review is scheduled around March to avoid redoing it later, followed by the introduction, discussions, conclusions, and time for reviewing and revising.

What travel and extra commitments are included only if conference acceptances happen?

Travel-linked work is treated as conditional. If a conference is accepted in February, it would be in Washington DC, with an added week of holidays. If IUI is accepted, it would be in Sydney at the end of March, again with extra holiday time and a likely weekend trip around Easter. Another planned break is a week early June, timed to avoid drama school and lecturing responsibilities. Later travel possibilities include Asia or Cape Town at the end of July, contingent on thesis submission and conference outcomes.

Review Questions

Which remaining thesis tasks are prioritized before the IUI deadline, and which tasks are explicitly deferred until after October 14?
How does the calendar integrate teaching/university weeks, business responsibilities, and personal events without losing month-level visibility?
What writing sequence is planned for the thesis, and why does reusing conference/journal work reduce later rework?

Key Points

1
Use a “year at a glance” page to map conference deadlines, teaching/university weeks, business dates, and personal events across the whole year, then add monthly detail for execution.
2
Treat the recommender-systems thesis as three connected deliverables: feature exploration, prediction modeling, and end-to-end recommender integration with UI validation and user evaluation.
3
Revisit the raw Strava-based dataset early enough to extract additional features and identify which training features correlate with marathon performance.
4
Use conference deadlines to drive research sequencing—schedule recommender-system/UI work for IUI and push feature engineering/prediction until after the IUI submission date.
5
Plan a late-stage “training descriptions” add-on as thesis-only work rather than a publishable deliverable.
6
Start thesis writing immediately after the January submission and target completion by June, with early chapters built from already-written conference/journal material.
7
Build in recovery time after submissions (e.g., a week near Halloween) and reserve conditional travel/extra weeks for conference acceptances.

Highlights

The calendar approach deliberately prioritizes a year-level overview to avoid losing the “shape” of upcoming months that happens when everything sits in a standard calendar grid.

IUI’s October 14 deadline forces a shift: feature engineering and marathon prediction are deferred, while the recommender system and UI validation become the main focus for that submission window.

Thesis writing is structured to minimize rework by starting with chapters tied to already-written conference/journal papers, then adding a full literature review around March.

User evaluation and UI validation are treated as essential components for strengthening the recommender-system thesis, not optional extras.

Travel plans are explicitly conditional on conference acceptances, with extra holiday time added only when those trips become real.

Topics

PhD Planning
Academic Calendar
Recommender Systems
Conference Deadlines
Thesis Writing Timeline

Mentioned

Ciara Feely
IUI
ICML
Strava