Inayah, Athi (2018) Analisis kemometrik menggunakan LDA (Linear Discriminant Analysis) dan PLS (Partial Least Square) dari sampel minyak babi dan minyak sawit berbasis data FTIR. Undergraduate thesis, Universitas Islam Negeri Maulana Malik Ibrahim.
|
Text (Fulltext)
14640001.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (4MB) | Preview |
Abstract
INDONESIA:
Minyak sawit merupakan kebutuhan pangan pokok masyarakat dan sering digunakan untuk mengolah bahan pangan, akan tetapi untuk menekan harga produksi suatu bahan pangan memungkinkan adanya penggunaan minyak sawit yang tidak murni, salah satunya bisa dicampur dengan minyak babi. Penelitian ini bertujuan untuk menganalisis perbedaan spektrum FTIR dari minyak sawit dan minyak babi serta campuran keduanya dengan konsentrasi yang berbeda dengan menggunakan analisis kemometrik berbasis LDA (Linear Discriminant Analysis) dan PLS (Partial Least Square). Spektrum spektroskopi Fourier Transform Infrared (FTIR) sebagai sensor yang belum mampu membedakan antara minyak babi dan minyak sawit serta campuran keduanya, sehingga diperlukan analisis kemometrik. Data LDA diolah dengan menggunakan metode stepwise untuk memilih variabel diskriminan yang memiliki kontribusi dalam fungsi diskriminan. Dari data LDA didapat fungsi diskriminan 1 dengan nilai varian sebesar 99.3% sedangkan fungsi diskriminan 2 memiliki kontribusi dengan nilai varian sebesar 0.07% yang artinya fungsi diskiminan 1 lebih dominan mampu membedakan antar kelompok minyak. Sedangkan analisis PLS sudah baik, hal tersebut dapat dilihat dari nilai R2 dan rendahnya nilai kesalahan validasi (RMSECV) dan kesalahan kalibrasi (RMSE).
ENGLISH:
Palm oil is the basic necessities of the community that often used to process food. In the other hand, the some producers often use impure palm oil to push down price production of foodstuff. Usually palm oil is mixed with lard. This research aims to analyze the difference of FTIR spectrum of palm oil and lard with mixing both of oils with different concentrations by using chemometrics based on analysis of LDA (Linear Discriminant Analysis) and PLS (Partial Least Square). Spectrum Fourier Transform Infrared Spectroscopy (FTIR) as sensor which is not able to distinguish between lard, palm oil and its mixture, so chemometric analysis is needed. LDA data are processed by using a method stepwise to select the discrimination variable which has a contribution in discrimination function. From data obtained of the discrimination function LDA 1 variant with a value amounting to 99.3% whereas discrimination function 2 has a huge contribution with the value of the variant of 0.07% which means that the discrimination function 1 more dominant able to distinguish between groups of oil. While, PLS analysis was already good, it can be seen from the value of R2 and low value validation errors (RMSECV) and the calibration error (RMSE).
Item Type: | Thesis (Undergraduate) | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Supervisor: | Tazi, Imam and Syarifah, Umaiyatus | |||||||||
Contributors: |
|
|||||||||
Keywords: | FTIR; Kemometrik; LDA; PLS; Chemometrics | |||||||||
Departement: | Fakultas Sains dan Teknologi > Jurusan Fisika | |||||||||
Depositing User: | Heni Kurnia Ningsih | |||||||||
Date Deposited: | 10 May 2019 14:58 | |||||||||
Last Modified: | 10 May 2019 14:58 | |||||||||
URI: | http://etheses.uin-malang.ac.id/id/eprint/14217 |
Downloads
Downloads per month over past year
Actions (login required)
![]() |
View Item |