Catatan Riki Akbar

Mengingat perkembangan akhir-akhir ini ketika kita banyak disuguhkan data dan visualisasi yang terdistorsi, ada semacam kewajiban moral dan intelektual untuk mengingatkan kita semua untuk tidak langsung tersilaukan oleh pernyataan-pernyataan yang megklaim merupakan produk dari analisis dan pengolahan data. Penggunaan terminologi yang tidak pada tempatnya (dan bahkan cenderung abusive ) seperti big data dan machine learning harus dikritisi habis-habisan, tidak peduli siapapun yang menggunakan istilah-istilah tersebut. Seringkali saya pribadi jumpai penggunaan tidak pada tempatnya atas istilah-istilah tersebut dilakukan oleh orang-orang yang bahkan tidak mengerti apa 'big data' atau 'machine learning' itu sendiri.  Belum lagi produk yang diklaim berasal dari so-called data analysis  seringkali digunakan sebagai justifikasi pembentukan opini publik atau bahkan lebih parah lagi:  policy . Kental dengan bias dan conflict   of interest .  Trust me, I've seen enough....

As I write this post, I have been waiting for the training process to be completed. It's still the first epoch but it feels like it is taking ages. Forgive me, by the time I uploaded this image, it has reached the 4th epoch.  The first epoch is yet to be concluded. The accuracy is somehow far from my expectation. Anyway, it's still on the first epoch, so things might be about to change. At this stage, I am surprised by how much a love-hate relationship is involved in training a classifier. You get excited when your result meets your expected threshold but most of the time, you are left disappointed.  But still, I cannot find any other things quite interesting to do. I love how data can transform into valuable insight. I love how to spot patterns out of so-called trash and raw data. But really, I am just too lazy to explore further and anticipate more options of workaround to improve the classifier's performance. Not to mention how intimidated I am to see the progress made b...

 

Let a string variable named s  = '!!ab#c???' Use the following lines to remove all punctuations from s The result of the code above: Clean and concise! London, January 26 2022

A data distribution (commonly presented using histogram) can be analyzed by utilizing these components S HAPE: Bell-shaped, Left-skewed, Right-skewed, etc. C ENTER: Mean and Median S PREAD: Maximum & Minimum value, Quartile, Interquartile range O UTLIERS: Analyze the existence of outliers. Visit this  to find out how to determine outliers . The relationship between mean and median on different shapes of data distribution typically can be considered as follows: Bell-shaped: Mean = Median Left-skewed: Mean < Median Right-skewed: Mean > Median The Five Numbers defines five statistical measures to obtain a dataset profile/summary so that the center and the spread of the dataset can be identified. The Five Numbers are: Min/Minimum Q1/First Quartile Median. Also known as Q2/Second Quartile Q3/Third Quartile Max/Maximum Occasionally, The Five Numbers are complemented by Standard Deviation and Interquartile Range to produce a more comprehensive preview of the dataset. Standard De...