Prediksi Kesiapan Sekolah Menggunakan Machine Learning Berbasis Kombinasi Adam dan Nesterov Momentum
Bibliographic record
Abstract
Kesiapan sekolah adalah aspek perkembangan anak yang berperan pada kemampuan anak untuk beradaptasi dalam sistematika pendidikan tingkat dasar. Berdasarkan Permendikbud, usia 7 tahun adalah usia yang tepat bagi anak masuk Sekolah Dasar, karena anak telah memiliki kesiapan fisik dan psikis untuk mengikuti proses pendidikan formal. Namun, setiap anak tidak memiliki kondisi yang sama pada usia tertentu. Sehingga, diperlukan Nijmeegse Schoolbekwaamheids Test (NST) untuk mengukur kesiapan sekolah. Instrumen NST hanya dapat digunakan oleh Biro Psikologi yang mempunyai kemampuan dalam melakukan asesmen psikologis. Sedangkan, guru serta orang tua yang memiliki peran dalam bentuk pemberian dukungan dan stimulasi pada anak tidak dapat menggunakan instrumen tersebut. Machine learning adalah teknik yang menggunakan algoritma untuk menemukan pola yang berguna dalam data. Berdasarkan data NST terdahulu, dapat dirancang model prediksi kesiapan sekolah yang akan memudahkan guru dan orang tua dalam mengetahui kesiapan anak untuk masuk Sekolah Dasar. Data penelitian adalah data administratif 225 siswa yang telah mengikuti tes kesiapan sekolah berbasis NST yang diselenggarakan oleh TK Ar-Rasyid pada tahun 2012-2018. Data administratif NST terdiri dari umur, jenis kelamin, urutan anak, jumlah saudara, status TK, pendidikan ayah, pendidikan ibu dan hasil kesiapan sekolah. Berdasarkan korelasi Chi-Square, variabel yang memiliki hubungan signifikan kuat terhadap hasil tes kesiapan sekolah adalah status TK, jumlah saudara dan umur anak dengan nilai p<.05. Penelitian menggunakan algoritma Artificial Neural Network dengan metode optimasi kombinasi Adam dan Nesterov Momentum. Pengujian menggunakan skenario 5-fold cross validation. Hasil penelitian menunjukkan bahwa kombinasi Adam dan Nesterov Momentum memiliki kinerja lebih baik daripada classical Adam dalam memprediksi kesiapan sekolah dengan akurasi 96% dan loss 0.06 dalam 1.98 seconds pada 10 neuron dan 1000 epochs. Abstract School readiness is an aspect of child development that plays a role in the child's ability to adapt in the systematics of elementary level education. Based on the Minister of Education and Culture, 7 years is the right age for children to enter elementary school, because children already have physical and psychological readiness to take part in the formal education process. However, every child does not have the same condition at a certain age. Thus, the Nijmeegse Schoolbekwaamheids Test (NST) is needed to measure school readiness. The NST instrument can only be used by the Psychology Bureau who has the ability to carry out psychological assessments. Meanwhile, teachers and parents who have a role in providing support and stimulation to children cannot use these instrument. Machine learning is a technique that uses algorithms to find useful patterns in data. Based on previous NST data, it can be designed as a school readiness prediction model that will facilitate teachers and parents in knowing the readiness of children to enter elementary school. Research data is administrative data of 225 students who have taken the NST-based school readiness test conducted by TK Ar-Rasyid in 2012-2018. NST administrative data consists of age, gender, child position, number of siblings, pre-elementary status, father education, mother education and school readiness results. Based on the Chi-Square correlation, variables that have a strong significant relationship to school readiness test results are pre-elementary status, number of siblings and age with p<.05. The research used Artificial Neural Network algorithms with a combination of Adam and Nesterov Momentum optimization method. Model testing used a 5-fold cross validation scenario. The results showed that the combination of Adam and Nesterov Momentum performed better than classical Adam in predicting school readiness with 96% accuracy and 0.06 loss in 1.98 seconds on 10 neurons and 1000 epochs.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.013 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".