Delivering python packages ? 🤕

I made an OpenAI GPT-3 powered chatbot web app with streamlit in Python. You can also try this service by getting an API key from OpenAI. They offer a free credit tier to try their some models and here I used text-davinci-003 that is the most capable model for NLP. 🤖❤

Did she tell that she invented or the page is saying ? If she has given proper credits to the resources she followed that is okay. This is a lil person who tries to build her career. Don't share much and disrespect publicly ❤️

Day 66 - #66DaysOfData Today I completed the last projects with the help of GitHub and achieved my @freeCodeCamp Scientific Computing with Python Certificate. 😊 #datascience #python #projects #computing

This is one of the best evidences for my life that only (hard + smart) work impacts forever 💜️ Thank you Coursera 😊❤️

I don't know who needs to hear this, but Time ≠ Money.

It's my balloons day 😊💫

A warm hug to all those who chase goals.

I'm so happy to share that I obtained my @IBM Machine Learning with Python certificate on @coursera. 👀✨ This is the 9 of 10 in the IBM Data Science Specialization. ❤️ #ibm #ibmdatascience #coursera #datascience #machinelearning #algorithms #python #share #content #university

90's kids are still thinking about how to find the correct partner to marry and balance work with life 😂

Day 98 - #66DaysOfData Today I completed the week 1 tasks in Data Analysis with Python IBM course 😊♥️ #python #dataanalysis

Me wondering Where should I go, Am I on the right path and am I capable of continuing. 😊❤️ #yesterday

I'm going to write about important topics related to Data Science in the year 2024, in a clearer and more interesting way for anyone interested in the Data Science field or already working within the industry. Let's dive into Machine Learning !! 1. What is Machine Learning (ML)? Simply put, Machine Learning is a subset of Artificial Intelligence (AI). This means that to understand AI, you should have a clear understanding and foundation in ML. In traditional programming, we provide the data and program, and computers give us outputs. However, in Machine Learning, we train algorithms using past data. Using new data, machines provide predictions about the future by identifying patterns. Normal programs can't predict the future by analyzing unseen data. Therefore, the main task is to create a ML model for the training process. 2. In which areas are ML actually used? We can observe a vast array of ML applications across every industry. ✅ Diagnosing Diseases ✅Recommendation Systems ✅ Sentiment Analysis ✅ ChatGPT, Grok like Chatbots ✅ Language Translation ✅ Self-driving cars like Tesla cars ✅ Preventing Cyberattacks ✅ Pest Detection 3. Types of Machine Learning Machine Learning also has its own categories. ✳️ Supervised Learning: If you have labeled data to train, this is for you. ✳️ Unsupervised Learning: Use this when you only have data without labels to discover underlying structures. ✳️ Reinforcement Learning: When you have no data. These should be described separately. 4. Is it simple to Build, Train, Deploy, and Monitor ML Models? Yes, if you have determination. You should learn to orchestrate many libraries in the Python language, different algorithms, gather quality data, and work with cloud technologies. 5. Are ML Models 100% accurate every time? Nope, Never. However, there are loss functions, optimization functions, data preprocessing, evaluation methods, tuning, and monitoring techniques to make ML models more useful and accurate. I hope that as a start, this writing refreshed your understanding of Machine Learning, and I invite you to delve deeper into complex topics in the future with me. #MachineLearning #Datascience #HappyNewYear

I'm so excited to share that today I got my @IBM Databases and SQL for Data Science with Python certificate on @coursera with Honors. 👀✨ This is the 6 of 10 in the IBM Data Science Specialization. ♥️ #ibm #ibmdatascience #cousera #datascience #sql #database #data #python

I updated my cover photo. 😊❤️

Am I the only one who didn't post a selfie recently here on X ? 😮❤️

Finally IBM Certified !! So This is the @IBM Data Science Professional certificate that I obtained by working past few months. 😊️❤️   I really appreciate my LinkedIn and Twitter connections who gave me at least one react when I share my daily updates with #66daysofdata. ❤️

Just wrapped up Day 1 on my trip to a very rural area in Sri Lanka. This was a long time dream but It's not easy to travel to these areas because of the hot weather. 🏝️😊❤️ 1. Thanthirimale 2. Baobab Tree in Mannar. ( One of the 3 for the whole country ) 3. Thalaimanner pier. #SriLanka #traveling

Day 7 - #66DaysOfData - Created and published a post related to Machine Learning. - Read 105 - 108 pages of the data science book ' Build a career in Data Science' by Emily Robinson and Jacqueline Nolis.

How much time do you spend on X for a day ? I normally spend 1-2 1/2 hrs in my active days on X.

My country view from the International Space Station. Thank you for this amazing photograph. 😊❤️🇱🇰

What is Supervised Learning? As I previously mentioned in my last post Supervised Learning is one type of Machine Learning. Here algorithms learn to find patterns and relationships from labeled training data. Simply, your algorithms learn to predict output variables using inputs ( Feature Variables). Let's dive deep !! We can divide Supervised Learning into two categories. 1. Regression - In these problems predicting continuous numerical values. Ex : Predicting house price based on features like number of bedrooms, number of bathrooms and area. 2. Classification - Here algorithms predict discrete labels. Ex : In Image recognition, predicts an image containing a cat or dog. There are many popular Supervised Learning algorithms. Mainly learning to implement them and the theoretical understanding of them is a minimum requirement for preparing Data science interviews. Here are some of them. ✳️ Linear Regression ✳️ Logistic Regression ✳️ Decision Trees ✳️ Random Forests ✳️ Gradient Boosted Trees ✳️ K-Nearest Neighbors ✳️ Support Vector Machines (SVM) ✳️ Neural Networks ✳️ Naive Bayes What are the advantages of using these Supervised learning algorithms to solve a problem ? ✅ ML model can easily predict with the help of prior experience. ✅ We know the exact classes. ex: for spam filtering class labels are spam and ham. What are the disadvantages of using Supervised Learning ? ✅ Supervised Learning is not for complex problems. ✅ We need prior knowledge about output classes. ✅ Depends on human labeled data. You can read more about this here : scikit-learn.org/stable/tuto… #DataScience #AI #MachineLearning

I wish you all a Merry Christmas 😊✨️

Day 11 - #66DaysOfData - Read 117 - 120 pages of the data science book ' Build a career in Data Science'. - Python for Everybody freeCodeCamp certification ☑️ Iterations : Loop Idioms ☑️ Iterations : More Patterns #datascience #machinelearning #python #career

Day 32 - #66DaysOfData Today I could complete another 4 parts related to Relational Databases of Python for Everybody freeCodeCamp Certification and do some python exercises. Believe Consistency is the key of any success. ♥️ #DataScience #Python

The people who are connected with me on X live in mostly USA and India. A few from Europe. Is that same to you ?

People who show fake kindness

Believe me or not. It feels like two years when you haven't lived with X two days. 😂

😊🖤 Today I achieved the @IBM Data Science Methodology Certificate on @Coursera. It mainly focuses what is methodology and why data scientists need methodologies also the six stages in the Cross Industry Process for Data Mining (CRISP-DM) methodology. 😊✨ #IBM #ibmdatascience

Happy Holi to all my Indian friends 😊🙌❤️

Seems like ChatGPT 3.5 knows about the one year celebration 😏❤️

Two years guys. I'm glad. I believe I have shared useful content related to Data science with you and has supported your works here too. Thank you for all care you have showed me in these two years. Love you all. 😊🔥 I'm not gonna stop sharing my knowledge and light with you. 😂😊✨️ #MyXAnniversary

Hey you, Stay positive 😊💫

Day 69 - #66DaysOfData Um working on Python Project for Data Science IBM course but a quite lazy day. ♥️

This is a concept I also believe always but actually speed matters in the current world. Dreams without deadlines always fail. When everyone measures your success according to your age and expects from you what others do at that age, it is difficult for a heart to deal.

Day 21 - #66DaysOfData - Going through pandas documentation and read some articles related to pandas to create my short note. - Read 139 - 142 pages of the data science book ' Build a career in Data Science' by Emily Robinson and Jacqueline Nolis. #datascience #pandas #python

Python Programmers 😎

Not AI generated 😂🥲❤️

Coding and software building are two separate things. 😏❤️

I love data science. Don't you?

Do you think that Safe Super intelligence is a good startup idea ?

My pinned post related to the chatbot development has got 100 likes today. I love that post very much. 😊❤️

Past two years I was always excited to share valuable resources and updates on AI and Data science with you. We shared, learnt, chat and laughed on X. I believe we had a great time and was supportive to each other to grow. Thank you for being a part of my journey too. Together, Let's explore more. 🙌😊❤️

Day 112 - #66DaysOfData Completed the tasks and obtained the certificate in Data Analysis with Python IBM course 😊♥️ #dataanalysis #python

Happy Valentine's Day to all !!

We should practice to be happy with what we already have, stay calm in the situations we don't get what we expected however if something really needed should work for it.

I'm really happy to tell you all that today I obtained my @IBM Data Visualization with Python certificate on @coursera. 👀✨ This is the 8 of 10 in the IBM Data Science Specialization. ♥️ #ibm #ibmdatascience #coursera #datascience #datavisualization #python #data #dashboard

Stay curious 😊💫

After 4 yrs of learning software engineering in university and doing self studies nearly one year in data science ( except ML DL modules in uni ) I figured out that multi-tasking is not gonna help. That leads to depression. If you want to be a master of something do one thing.

Today I received this valuable gift from @DataChaz 😊❤️ Web Application Development with @streamlit book by Mohammad Khorasani. 😍 Thank you so much @DataChaz for this help you gave me to learn further in data science.❤️ I'm so glad to have a friend on twitter like you ✨️

Do you believe that if you do good to others it will come back to you in unexpected ways ? Ex : food, support for learning

Happy to be here on X 😊🔥

I love potato chips since childhood. What is the most favourite thing you love to buy in your entire life ?

What is Bayesian Learning ? It's also a type of Machine Learning. Bayesian Learning is a more statistical approach where models are constructed using probabilistic principles and Bayes' theorem to predict and update beliefs. It's like your grandfather telling you about tomorrow's weather by looking at the sky. He knows in this period of time what happened earlier and what was like the previous days and what will be the next. That means it's based on prior knowledge and new evidence. To learn Bayesian learning firstly we should understand Bayes' theorem. Bayes theorem Bayes theorem simply describes the probability of an event based on prior knowledge of conditions that might be related to that event. P(A | B) = P(B | A) * P(A) / P(B) P(A|B) - When the event B is already occurred, Probability of event A P(B|A) - When the event A is already occurred, Probability of event B P(A) - Probability of event A without giving any condition. P(B) - Probability of event B without giving any condition. What are the advantages of using Bayesian Learning? ✳️ When you have limited data you can use your prior knowledge. ✳️ Because of the probabilities updating nature you can create responsive models. ✳️ Because of the nature of handling uncertainty, one can always be aware of the uncertainty in prediction results. What are the Disadvantages of using Bayesian Learning? ✳️ For complex models it may require heavy computational power. ✳️ Can be biased and incorrect because of using prior knowledge of a person or a group of persons. ✳️ Sometimes interpreting the inputs and results can be complex. Famous Bayesian Learning Models ✳️ Naive Bayes Classifier ✳️ Bayesian Linear Regression ✳️ Bayesian Networks Where can we find the applications of Bayesian Learning? ✳️ Spam Filtering ✳️ Sentiment Analysis ✳️ Assisting in Disease Diagnosis ✳️ Recommendation Systems ✳️ Robotics and Autonomous systems So the next time with this knowledge can you guess what will happen based on what you have already done or your best friend has done ? Thanks for Reading !! #DataScience #MachineLearning

The Google interview warmup tool is a quick and a best way to prepare for facing interviews. You can practice answering interview questions, get insights detected by Machine Learning and can improve. Preparing for the interviews is hard when you don't have friends, family or mentors who can help you practice and prepare in a different field. But you should practice because interview facing is also a skill to master to land at a job. If you couldn't have that skill your skills to complete the job is going hidden to the world. If there are any other effective ways to practice interview questions please let me know in the comment section. That'll be helpful !!

Happy easter to all 😊❤️

One of my best friends told me about his preference to drink energy drinks and low alcohol level cold beers. How it started. In his age at 16-18 he has seen one of his tution teachers using an energy drink in the intervals to get some refreshment. The sound of opening a can was super cool. After, in his early twenties he has seen in a movie scene people got some beer bottles as a win. That was a goosebumps moment to him. I can't categorize drinking them as a good or bad habit exactly. But I understood people get influenced by what they see as some winning and giving a refreshed mind. What we can get from here ? I suggest you to follow the people who succeeded in your niche and experience how they are achieving and winning. Actually it can give you the best habits without you knowing to support you to climb the success ladder.

You're capable and you know it. Embrace it !!

AI generated or not ?

Hi! I'm really excited to share with you all my today's achievement. I obtained my @IBM certificate 'Python for Data Science, AI & Development' on @coursera. 😊😊 I learned here Python basics.♥️ #ibm #ibmdatascience #coursera #datascience #datastructures #ai #development #python

What is Unsupervised Learning ? Unsupervised Learning means you train your Machine Learning Models with unlabeled data to find hidden patterns, structures or relationships. You don't know the target values. That means Unsupervised Learning is really great for the exploration purposes. Where can we see applying Unsupervised Learning techniques? ✳️ Customer Segmentation ✳️ Anomaly Detection ✳️ Dimensionality Reduction in the EDA process ✳️ Recommendation Systems ✳️ Target marketing There are different Unsupervised Learning Types. 1. Clustering Clustering groups similar data points together. That means find similarities. Ex: You group your blue sock with a blue sock based on the color. Red with a red. Movies grouping according to genre is a similar situation. 2. Association Rules Association rules means you find relationships. Ex: Think about shopping habits. If someone buys a pizza they buy a coke too. If someone buys a phone they buy a charging adapter/power bank too. This helps businesses to place the products in the right places in a shop or do target marketing. 3. Dimensionality Reduction This technique is used to transform data from high dimensional spaces to low dimensional spaces without losing too much detail from original data. Ex: This post is kind an example. I'm simplifying and writing information about unsupervised learning without losing the meaning of unsupervised learning. There can be full books with hundreds of pages and coding examples on this topic. But reading this post you can have a good visualization about the concept in more efficiently. What are the famous Unsupervised Learning algorithms? ✳️ K-Means Clustering ✳️ Hierarchical Clustering ✳️ DBSCAN ✳️ PCA (Principal Component Analysis) ✳️ Apriori Algorithm ✳️ t-SNE (t-Distributed Stochastic Neighbor Embedding) ✳️ Autoencoders ✳️ Gaussian Mixture Models (GMM) There may be a best recommendation system for X. That's why you found me with your similar interests. Thanks for Reading !! #DataScience #AI #MachineLearning

I know you're really busy You do coding You have many meetings You're learning But remember to care about your loved ones too. Give a few mins !! For the people who love you it may feel like a year. At the end without good bonds what are you truly achieving ? And give a few minutes of relaxation to yourself too.

ChuanhuChatGPT 😊🔥 A beautiful GUI for ChatGPT API and many LLMs. Supports agents, file-based QA, GPT finetuning and query with web search. 👇😊❤️

It's Happy New Year Day in Sri Lanka 😊❤️ This new year starts at 9:05 pm.

I'm so excited to share that today I obtained my @IBM Data Analysis with Python certificate on @coursera. 👀✨ This is the 7 of 10 in the IBM Data Science Specialization. ❤️ #ibm #ibmdatascience #coursera #datascience #dataanalysis #python #statistics #data #visualization

I was not very active in the past few days on X because of the trip. Now I'm here. 🌚❤️ I'm available again to learn data science, share updates, connect and distract you. 😂 Now you are not alone on X ma friends. 🌚🤗❤️

Happy Diwali to all my Indian friends who celebrate this day if I missed to wish you in your personal posts. Have a bright future as this 🪔. With Love Roshini ( Light ) 😊❤️

10K posts !! That was a fun and useful journey with you all 😊❤️🔥

Missed the days on this app people shared more technical content. The days people share about kaggle competitions much, about statistical models and everything.

X is the best platform that ever happened to me. I learnt a lot from here not only related to data science but also about life and connections.

Great Sachin ❤️

The only hope I have for success in a data science role now is me. My soul. Roshini !!

3rd and the Final day of my trip to Mannar and Jaffna 😊🌿 We went to Nagadeepa viharaya and kovil, General Denzil Kobbekaduwa Memorial place, Jaffna old fort, Sri Nagavihara, Nallur kovil, Rio ice cream, Kadurugoda viharaya and more. I carefully selected these few photos to share here from hundreds. I love those. Real GOAT is there. Will messi agree with me? 🤭❤️ Also me and my whole time best friend my amma. 😊❤️ As a summary it was very harsh but worth the trip. I came again to western province with a lot of happy memories. Farwell Jaffna 😊❤️ I should thank my X friends for admiring my journey these days. That means a lot to me. Thank you all 😊❤️ #SriLanka #traveling

When life seems to be a jumbled equation, simplify it by subtracting the unnecessary elements. 🥹❤️

When I see New LLM model releases.

On Day 2 of our trip we came to Jaffna from Mannar. For my life time as a Sinhala buddhist I never thought I could visit Jaffna here because of the past war time period. Now everything is okay and I'm so happy that Jaffna is developing. Really peaceful. There are many shops normally in the city. I hope education and technology will also improve in Jaffna to have more opportunities for the next generations. Sri lankan tamil hindu culture is a very rich one. Jaffna we can see there are many kovils. Food is awesome. I think as normal people we can do as a support to develop is visit Jaffna and admire and experience here more. We visited Delft island and had all the experiences there including another baobab tree visit, watching wild horses and old ruins. The boat ride was just awesome to the delft island. 😍🔥 I think my inner child played really well in there. 😂 And look at me with that red thing in my forehead from kovil. I love that 🤭❤️ I'm blessed. 😊❤️ #SriLanka #traveling

I'm working on my new notebook on Feature Scaling to explore more about Feature Engineering. 😊🔥 I'm happy because I could learn about different scalers practically from kaggle notebooks and other resources using sklearn.preprocessing module. Should continue 😊❤️ #DataScience #FeatureEngineering

There's only one thing important in our lives. That is, DATA.

You know exactly what your potential is. But you're not taking actions. Why is that babe ?

Who clicked this ? 😂🔥

When they ask about your passions. Some people get jealous of you for your goals. Don't ever say fully what you need in career life to others.

Give time. Be patient about yourself. As you do for others, Yourself needs love and care too.

Today is my Day 66 - #66DaysOfData In these 66 days I was able to continuously learn and share my experiences because of I took this challenge #66DaysOfData. I should truly appreciate @KenJee_DS for initiating this.♥️ I'm glad to make a brief summary of What I did these days

Our mindset is everything. We all here are not from rich families. Normally children from rich families get better resources and opportunities than poor children from the start of their montessori time. Poor children get skills gaps. The most crucial asset the rich children get is should try mindset more than any other resources they get. They are not afraid of experimenting with devices or materials they already have. Poor children afaid to break the things. They are afraid to disappoint parents. But we all come to an age where we can realize what we should do to improve our mindset to have a let me try this mindset. Search, Read and Experiment more. Anyone can start and fill the skills gaps. Don't die as a poor person.

One of my best friends asked me some scenary kind of photos from my trip to Jaffna aka YALPANAM. He really loved to give a simple edit and gifted me those photos to upload in my X profile. That's what he only wanted. He added my name as Roshinifer to photos. Thank you !! 😊❤️

Thank you so much Kris (@icreatelife) for conducting the awesome "Create Your Worlds and Characters with AI." session to teach us about generating AI arts with Adobe Firefly. I learnt a lot and that was not boring at all. I hope you'll love this art about your cute Pablo. Thank you again !! 🙌😊❤️

I'm really excited to say you all that Today I achieved my Tools for Data Science @IBM Certificate on @coursera. 😊❤ It mainly covers working in Jupyter Notebooks on IBM Watson Studio, GitHub, RStudio, IBM SPSS Modeler and SPSS Statistics. #ibm #ibmdatascience #coursera

I'm Devin and friendly. Don't afraid mate !!

What is Data Labeling (Data Annotation)? Data labeling is a process of identifying raw data (texts, images, videos) and giving a meaningful label or multiple labels to them to use in training and testing Machine Learning models. Basically it’s a Human In The Loop(HITL) process. For a true Well-labeled dataset, we call "ground truth" in Machine Learning. If your ground truth is accurate, your model's accuracy will be better. Therefore it's crucial to give time and resources to ensure highly accurate Data Labeling. Where do we need labeled data? In Computer vision for object detection, identify key points in an image, image segmentation and for classification need labeled image datasets. In Natural Language Processing for Optical character recognition, Entity name recognition and sentiment analysis need labeled text data. For audio processing and video processing samely need labeling. There are different ways of Data Labeling. ✳️ Internal Labeling - Data science experts within the organization involved in this. ✳️ Synthetic Labeling - Generate new data from pre-existing data using machines. ✳️ Programmatic Labeling - Automated but need to do QA. ✳️ Outsourcing - Take the service from a third party. ✳️ Crowdsourcing - Ex : Recaptcha Though there are benefits like better usability and accuracy of data still have challenges to do labeling. ✳️ Prone to human errors. ✳️ Expensive. ✳️ Time consuming. However there are ways to do labeling with increasing efficiency. ✳️ Intuitive and streamlined task interfaces - Helps human annotators to have productivity. ✳️ Labeler consensus - Taking multiple labelers responses to have the finals. ✳️ Label auditing - Verify and update. ✳️ Active learning - Model itself participates in the labeling process. Incorrect and biased labeling can significantly impact the performance of Machine Learning Models. Therefore If you know the current best ways/tools to do labeling to create datasets efficiently please state here. Thanks for Reading !!

Omg Look who just followed me. 🥺🔥 You all know about @MindsDB right ? Thank you so much 🥺🙌

What is Semi-supervised Learning? This is an approach that uses labeled data and unlabeled data both. As you learnt from my previous post about Data Labeling, it's really expensive and very time consuming to use labeled data in a huge amount. In that case, a semi-supervised learning approach is a solution to that using small amounts of labeled data and large amounts of unlabeled data. Labeled data helps the model to learn to predict the labels for new data while we use unlabeled data to refine the model and its performance. It increases the generalization ability of the model. There are different semi-supervised approaches. ✳️ Self training - In here the model learns from labeled data then used to label unlabeled data. ✳️ Co-training - Here two models train on different labeled data parts and after each model use to label unlabeled data. ✳️ Multi view training - This is a variant of Co-training. It’s very useful when you have different types of data. ✳️ SSL using Graph Models - When the data has a complex structure, this is useful to represent them as a Graph. What are the advantages of using semi supervised learning? ✳️ Prediction and exploration both can be achieved. ✳️ Cost optimization for labeling. ✳️ Improved generalization. ✳️ Handling rare cases. ✳️ Flexibility. ✳️ Robustness. Though the semi supervised learning has these advantages there are limitations like computational complexity, difficulty of choosing a model architecture and noise data. Where can we use semi supervised learning ? ✳️ Optical Character Recognition ✳️ Face Recognition ✳️ Speech Recognition ✳️ Text Classification ✳️ Image Classification ✳️ Recommender systems ✳️ Anomaly Detection ✳️ Drug Discovery and Bioinformatics ✳️ Internet Content Classification Thanks for Reading !! #MachineLearning #DataScience

Please support this man. He is sharing much valuable Machine Learning related contents. Not only theories. He is sharing notebook practical examples too. 👇😊❤️

Some people never value long paragraphs until they never hear from that lovely person who sent them.

How amazing is this @CocaCola poster ? 🥹❤️ I got a instant thirst when I saw this. I generated this using Bard. 🥹❤️

Had a very valuable read about building an AI powered chatbot with OpenAI and RAG. Check his medium blog 👇😊🔥

25 X friends have said that Deep Learning is the most sexist. 😊🔥

When I got the chance to comment on my X friends posts, it is an another level of happiness. 😊❤️

Change is natural. So that should be easy for us to do on ourselves too.

I generated this amazing New York City (NYC) image through Bard. 😍 That's so much good in deatails than I expected. Bard is using @Google 's Imagen 2 model. What will happen when the Gemini Ultra came ? 😊🔥

Forgive your parents. Forgive ex lovers. Forgive friends. Forgive others. Forgive everyone. Most importantly forgive yourself.