代写EG2002、代做Java/Python程序语言
EG2002 – Coursework
Analysis of a 3D printing dataset
Description:
In recent decades, 3D printing has emerged as a disruptive manufacturing technique,
offering unparalleled flexibility in material selection and the ability to produce complex
geometries that were previously difficult or impossible to achieve.
Figure 1. 3D printer.
The quality of 3D printed pieces is influenced by various parameters such as print speed,
temperature, and material type, which directly impact the material's performance and
finish. In this project, you will explore data obtained from pieces printed using different
parameters and develop models to predict tensile strength, elongation, and surface
roughness.
The dataset contains the following 9 features:
Feature Type Description
Layer Height (mm) numerical Thickness of each layer of material deposited during printing.
Wall Thickness
(mm)
numerical Thickness of the solid walls of the object.
Infill Density (%) numerical Percentage of the object's interior filled with material.
Infill Pattern () ordinal The geometric pattern used to fill the interior of the object.
Nozzle Temperature
(Cº)
numerical Temperature of the material exiting the printer nozzle.
Bed Temperature
(Cº)
numerical Temperature of the printer bed where the object is laid down.
Print Speed (mm/s) numerical Speed at which the printer nozzle travels while printing.
Material () nominal The filament or material used for printing the object.
Fan Speed (%) numerical Speed of the fan cooling the printed object. Affects how quickly the
material cools and solidifies.
Figure 2. Infill Pattern and Density.
The dataset contains the following 3 labels:
Labels/Targets Type Description
Roughness (µm) numerical Indicates the surface texture’s irregularities.
Tensile Strength (MPa) numerical Strength of the object.
Elongation in % numerical Indicates how much the object can stretch before breaking.
Tasks:
For this coursework, you will need to create a notebook to read the dataset
“data_3dprinter.csv”, assess and clean the dataset and visualise the data. Please include
all necessary explanations and conclusions as markdown cells in the notebook.
Select and train the most suitable models (you may use more than one) to predict
roughness, tensile strength, and elongation. Evaluate and compare the performance of the
models using appropriate metrics. Please include all necessary explanations and
conclusions as markdown cells in the notebook. Create a separate notebook for each type
of model considered.
Compare the results obtained using algorithms from libraries (e.g., scikit learn) and your
own implementation.
Additional information:
Split your data into a training set containing 80% of the data and a testing set containing
the remaining 20%. Tune the hyperparameters of the models to achieve the best results.
Explore and discuss in the notebook whether the models may be overfitting.
Include all your notebooks in a single zip file named “Group Number.zip” and upload it to
Moodle.