No abstract

No abstract

The use of machine learning harbours the promise of more accurate, unbiased future predictions than human beings on their own can ever be capable of. However, because existing data sets are always utilized, these calculations are extrapolations of the past and serve to reproduce prejudices embedded in the data. In turn, machine learning prediction result raises ethical and moral dilemmas. As mirrors of society, algorithms show the status quo, reinforce errors, and are subject to targeted influences – for good and the bad. This phenomenon makes machine learning viewed as pseudoscience. Besides the limitations, injustices, and oracle-like nature of these technologies, there are also questions about the nature of the opportunities and possibilities they offer. This article aims to discuss whether machine learning in biomedical research falls into pseudoscience based on Popper and Kuhn's perspective and four theories of truth using three study cases. The discussion result explains several conditions that must be fulfilled so that machine learning in biomedical does not fall into pseudoscience

People with speech disorders could have social and welfare difficulties. Therefore, the silent speech interface (SSI) is needed to help them communicate. This interface decodes the speech from a human's biosignal. The brain signals contain information from speech production to cover people with numerous speech disorders. Brain signals can be acquired noninvasively by electroencephalograph (EEG) and later transformed into the features for the input of speech pattern recognition. This review discusses the advancement of EEGbased SSI research and its current challenges. It mainly discussed the acquisition protocol, spectral-spatial-temporal characterization of EEG-based imagined speech, classification techniques with leave-one-subject or session-out cross-validation, and related real-world environmental issues. It aims to aid future imagined speech decoding research in exploring the proper methods to overcome the problems.

Penelitian ini bertujuan untuk mengetahui pengaruh likuiditas, ukuran perusahaan, dan kepemilikan saham publik terhadap pengungkapan laporan keuangan pada perusahaan properti & real estat yang terdaftar Bursa Efek Indonesia periode 2017-2019. Penelitian ini menggunakan metode penelitian kuantitatif dengan data sekunder berupa laporan keuangan tahunan yang diperoleh dari laman Bursa Efek Indonesia atau website perusahaan. Penelitian ini terdiri dari total sampel 95 laporan tahunan perusahaan properti & real estat. Teknik pengambilan sampel yang digunakan yaitu purposive sampling. Teknik analisis yang digunakan dalam penelitian ini yaitu regresi linear berganda dengan alat uji SPSS versi 25. Hasil penelitian ini menunjukkan bahwa (1) Likuiditas tidak berpengaruh terhadap pengungkapan laporan keuangan, (2) Ukuran perusahaan berpengaruh positif terhadap pengungkapan laporan keuangan, dan (3) Kepemilikan saham publik tidak berpengaruh terhadap pengungkapan laporan keuangan.Kata Kunci: Likuiditas, Ukuran Perusahaan, Kepemilikan Saham Publik, Pengungkapan Laporan Keuangan.