tu delft phd thesis repository

Guidance on data sharing for current and new doctoral candidates

All doctoral candidates who started their PhD at TU Delft on or after 1 January 2019, are required to publish all research data and code supporting the results described in their PhD thesis in 4TU.Centre for Research Data (or another suitable data archive) before graduating.

All doctoral candidates who started earlier than in January 2019, are strongly encouraged to publish their research data.

Why do I need to share my research data?

As a doctoral candidate at TU Delft you are bound by the Netherlands Code of Conduct for Academic Practice . The Code requires that research data needed to validate scientific findings are available. In addition, making data Findable, Accessible, Interoperable and Re-usable (FAIR) is also a requirement of the TU Delft Research Data Framework Policy .

But, most importantly, sharing your data has plenty of benefits.

Benefits of Data Sharing

Sharing your research data helps moving human knowledge forward. Others, instead of re-inventing the wheel and repeating your research, can simply re-use and cite your data (and you are more likely to gain new collaborators).

Sharing your data also makes all your findings more reproducible and transparent. In addition, your thesis (and any research papers resulting from your data) will be more convincing (also to your examiners and reviewers).

Finally, there are multiple (selfish!) benefits of data sharing. Sharing your data with 4TU.Centre for Research Data helps maximise visibility and impact. All research data that is published in the data archive receive a Digital Object Identifier (DOI). This identifier defines a persistent link so that the research data can be cited by others. Published data deposits are also visible to search engines, such as Google.

And finally, in the future you will be always able to find and access all your data files - no more searching through multiple hard drives or thinking ‘where did I put it’... Most likely you will be the first re-user of your own data.

To summarise, by sharing your data you:

• Get the deserved credit for all your hard work => you can add the DOI to your data in your CV for your next job application
• Make your research findings more reproducible and trustworthy
• Allow others re-use your data, build up on your findings and create more opportunities for collaboration
• Increase your visibility and impact
• Can easily find your own data files, years after graduation

Want to know more about maximising visibility and impact of your data? Have a look at the dedicated course Sharing your Research With the World or read the excellent blog post published by Nature .

What do I need to do now?

Wherever you are on your PhD journey, remember to manage your data properly. Managing your data properly will save you a lot of time at the end of your PhD, when you need to gather all your files to prepare them for upload to the data archive.

TU Delft Library prepared a comprehensive website with guidance on research data management throughout the research lifecycle: https://www.tudelft.nl/library/research-data-management

And if you need any support with data management, there is a dedicated Data Steward at your faculty , who will be happy to help you.

Share this page:

• Research repository
• Education Repository
• Cultural Heritage

Prediction of Aircraft Take-off Weight using Machine Learning

Gheorghe, Andrada (TU Delft Aerospace Engineering)

Delft University of Technology

Aerospace Engineering

Predicting aircraft Take-Off Weight (TOW) has been a long-sought task by aviation stakeholders, especially for operational and regulatory bodies involved in flight planning. Unfortunately, TOW being a sensitive parameter to operational trends and cost indices, aircraft operators tend to keep it confidential. In recent years, Machine Learning (ML) algorithms have achieved increased prediction accuracy and capabilities in the field, provided the availability of TOW data. This paper studies the implementation of gradient boosting algorithms as well as Random Forests to better understand which algorithm is best-suited for aircraft TOW prediction (prior to take-off) solely based on Flight PLan (FPL) and Terminal Aerodrome Forecast (TAF) parameters. The study focused on flights at Amsterdam Airport Schiphol (AMS) for training the algorithms, using an 80-20% train-test split. Between Gradient Boosting Decision Trees (GBDTs), LightGBM, XGBoost, and Random Forests, GBDTs achieved the smallest Mean Absolute Percentage Error (MAPE) with 1.71 and 2.17% on the training and testing datasets, respectively. The most influencing feature proved to be the requested cruise speed, followed by great circle distance between airports, and aircraft type. The model was validated on Paris - Charles de Gaulle Airport (CDG) and Brussels South Charleroi Airport (CRL), proving its independence from airport type. However, the distribution of flights in the training dataset, especially that of aircraft and airline types, proved to be an influencing factor for the model's applicability to other airports. Future work includes expanding the training dataset to all flights in the European network, and introducing trajectory-based features such as aircraft speed intent. With a larger training dataset, neural network algorithms could also be explored. Finally, regarding the improvement of trajectory predictions, it was found that better accuracy of TOW predictions does not suffice and that other operational parameters' effect should be investigated, especially speed profiles.

Take-off Weight Machine Learning Gradient Boosting Random Forests Flight Planning Air Traffic Management

http://resolver.tudelft.nl/uuid:3a2b3d2e-e4f3-4042-9172-f03ccc67cda1

50.879640,4.432340

Student theses

master thesis

© 2024 Andrada Gheorghe

• Help & FAQ

View Scopus Profile

• Civil Engineering & Geosciences , Rivers, Ports, Waterways and Dredging Engineering

Fingerprint

• 1 Similar Profiles
• data INIS 100%
• Model Earth and Planetary Sciences 96%
• coastal regions INIS 63%
• water INIS 62%
• dynamics INIS 57%
• management INIS 49%
• simulation INIS 49%
• rivers INIS 45%

Collaborations and top research areas from the last five years

Dive into details.

Select a country/territory to view shared publications and projects

Research output

• 11 Conference contribution
• 7 Other contribution
• 3 Comment/Letter to the editor
• 1 Conference article
• 1 Web publication/site
• 1 Dissertation (TU Delft)

Research output per year

Using an idealized network model as the physical module for a salt intrusion serious game

Research output : Contribution to conference › Abstract › Scientific

• intrusion 100%
• Intrusion 100%
• Module 100%

Corridor Scale Planning of Bunker Infrastructure for Zero-Emission Energy Sources in Inland Waterway Transport

Research output : Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific

• emission 100%
• inland waterways 100%
• transport 100%
• energy sources 100%
• planning 100%

Drivers and barriers to knowledge exchange through an envisioned online platform for transdisciplinary research projects

Research output : Contribution to journal › Article › Scientific › peer-review

• barriers 100%
• Transdisciplinary 100%
• Online Platform 100%
• Research Project 100%
• Knowledge Exchange 100%

Forecasting estuarine salt intrusion in the Rhine-Meuse delta using an LSTM model

• forecasting 100%
• rhine river 100%
• Memory 100%

Simulating for sustainability: Alternative operating strategies for energy efficiency

Research output : Contribution to journal › Article › Professional

• energy efficiency 100%
• simulation 100%
• sustainability 100%
• Energy Conservation 100%
• Sustainable Development 100%

OpenCLSim (Version 1.0.0)

van Koningsveld, M. (Creator), den Uijl, J. A. W. (Creator), Baart, F. (Creator) & Hommelberg, A. (Creator), Zenodo, 20 Jul 2019

DOI : 10.5281/zenodo.3251546 , https://github.com/TUDelft-CITG/OpenCLSim/releases/tag/1.0.0

Dataset/Software : Software

OpenCLSim (Version 0.3.0)

van Koningsveld, M. (Creator), den Uijl, J. A. W. (Creator), Baart, F. (Creator) & Hommelberg, A. (Creator), Zenodo, 10 Jul 2019

DOI : 10.5281/zenodo.3304278 , https://zenodo.org/record/3304278#.YH2m6D-xV1M and one more link , https://github.com/TUDelft-CITG/OpenCLSim/tree/0.3.0 (show fewer)

Press/Media

How it would have been possible for the three alcatraz escapees to survive the infamous prison break.

Rolf Hut & Fedor Baart

1 Media contribution

Press/Media : Public Engagement

Delftse wetenschappers ontsnappen van Alcatraz

Olivier Hoes , Rolf Hut & Fedor Baart

22/03/16 → 26/03/16

4 items of Media coverage

'Rond middernacht verandert de stroming in de baai'. Door Mark Hamer

'ontsnappers alcatraz mogelijk levend aan land''. door redactie, actueel: ontsnapping uit alcatraz. jeroen dirks interviewt rof hut.

Navigation Menu

Search code, repositories, users, issues, pull requests..., provide feedback.

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly.

To see all available qualifiers, see our documentation .

• Notifications

A LaTeX template for creating beautiful PhD theses, originally created by TU Delft.

Inventitech/phd-thesis-template

Folders and files, repository files navigation, improved tu delft phd thesis template.

This is an improved version of the TU Delft PhD thesis template . It features a large number of changes to increase both on- and off-screen readability and quality, as well as reduce printing costs. It is a double-sided, colored dissertation style with hyperlinks (although many of these parameters can be easily changed).

How to Pay Attribution

I did not create the original template. However, I would appreciate it if you did keep the note that you are using this adopted version of the template. Moreover, would you send me a message (moritzbeller -AT- gmx -DOT- de) or create an issue/PR that you are using this style, together with your expected defense date, university and research group?

Theses Using This Style

• Moritz Beller , 2018-11-23, Software Engineering Research Group, TU Delft
• Anand Sawant , 2019-10-10, Software Engineering Research Group, TU Delft
• Davide Spadini, Software Engineering Research Group, 2021, TU Delft
• Vladimir Kovalenko, Software Engineering Research Group, 2021, TU Delft
• Alex Salazar, Computer Science Intelligent Systems, TU Delft

Here is a side-by-side comparison of the two styles.

To get a full picture of how this style looks, have a look at my PhD thesis , which is set using the defaults in this repository. Note that there are small differences between the online and print versions (such as cutter's marks to trim pages and page alignment on the print version). These are hard to get right, and I optimized them using two physical proof prints of my thesis.

The layout in this repository is very well-readable, but rather close to the low limits for font size and page utilization that I would use (i.e., if anything, I would recommend to increase font size, line height or page margins).

Key Improvements

Specifically, the changes in this template make these key improvements over the TU Delft dissertation house style: It ...

• actually compiles (see CI).
• has been used successfully in my dissertation and been approved by the TU Delft Graduate School.
• makes the page layout much better suited to printing, because of an increased binding offset width. Text does not go into the binding fold of the book. The writable part of the page is also centered better when printed.
• uses much nicer fonts for on- and off-screen readability, namely Libertinus for regular text and Inconsolata for mono-spaced elements such as source code listings.
• comes with a large list of already included packages, placed in a compatible order (some LaTeX package orders cause hard-to-debug compile errors).
• reduces the number of pages that need color, thus reducing printing costs, while pertaining TU Delft's signature blue for key elements such as chapter titles.
• makes for more crisp printing, by converting 90% gray anthracite to a fully black tone. Gray scale values between 30% and 90% typically look blurry or faint in print because the printer simply uses less ink for them. To cater for this, I reduced the header text size to not interfere with the running text below.
• keeps a consistent style of reporting page numbers on top of pages.
• makes the screen and print versions of the PDF resemble each other more, while keeping the respective benefits of each format (for example, for on-screen PDFs, no offset binding adjustment is needed).
• makes the propositions page now fit a bigger amount of text. Moreover, the format of the propositions page now fits inside the thesis, instead of being the same size (which would stand out!).
• changes a couple of questionable defaults, such as the copyright notice by the author.
• uses glyphs and symbols to refer to papers, and gives presets for the CV, list of publications and the IPA dissertation series.
• includes commands to generate ready-to-be-printed PDFs with all fonts embedded, a typical requirement to publish PDFs.
• is also highly recommend to use the LaTeX URL Eternalizer .
• comes with a useful set of predefined commands, such as circled , ahref and useful predefines for lstlisting and such.

Setup and Installation

See the original README .

There are three document options to the provided dissertation.cls style -- be sure to use print when sending to the printer.

Mac users will have to install the Libertinus and Inconsolata fonts on their systems.

This template is based on commit ff9d073 of TU Delft's template (dissertation template from July 2nd, 2015, strangely referred to in the TeX document as '2013/07/08 v1.0 TU Delft dissertation class'), which was the latest available on 13-11-2018. Unfortunately, I did not have access to their repository.

• Makefile 0.7%