Machine Learning Jobs

190 were found based on your criteria {{|number:0 }} were found based on your criteria

show all
  • Hourly ({{ jobTypeController.getFacetCount("0")|number:0}})
  • Fixed Price ({{ jobTypeController.getFacetCount("1")|number:0}})
Hourly - Expert ($$$) - Est. Time: 3 to 6 months, Less than 10 hrs/week - Posted
Hi I’m looking for some one that i can knows how to build bots for marketing and shopping with natural language and AI please make sure you have some examples before submitting my main focus would be is to get on to the facebook messenger and IOS10 platform first This will be an full time ongoing high paid gig if I feel we have a great working chemistry after a few months​ coding and assisting in a/b split testing AI bots​ natural language, AI, Coding experience and facebook messenger
Skills: Machine learning Artificial Intelligence ICD Coding MySQL Programming
Fixed-Price - Intermediate ($$) - Est. Budget: $50 - Posted
Chatbot Project Knowing what to say is not always easy - especially if you're a chatbot. Generating answers from scratch is very difficult and would most likely result in nonsense or worse - but definitely not a pleasant user experience. Therefore we're taking one step back and instead provide the correct replies which now "only" have to be chosen in the right dialog context. In this project you're given a dataset with fictional dialogs (adapted from [1]) from which one reply is missing and additionally a list with all missing replies. Your task is to map all missing replies to the correct conversation. Dataset The dataset consists of 4 files: train_dialog.txt and test_dialog.txt each contain the conversations. The format is always c##### indicating the conversation number separated by +++$+++ from the reply text. For example one conversation from the training set is the following: c03253 +++$+++ Wow! This is like my dream room! Are these all records! c03253 +++$+++ I have about fifteen hundred 78s at this point. I've tried to pare down my collection to the essential... c03253 +++$+++ God, look at this poster! I can't believe this room! You're the luckiest guy in the world! I'd kill to have stuff like this! c03253 +++$+++ Please... go ahead and kill me! This stuff doesn't make you happy, believe me. c03253 +++$+++ You think it's healthy to obsessively collect things? You can't connect with other people so you fill your life with stuff... I'm just like all the rest of these pathetic collector losers. All original conversations are at least four lines long and always the second to last line is missing in the dialogs. The missing replies are found in the files train_missing.txt and test_missing.txt respectively. For the training dialogs, the conversation number is given with the reply as in the dialog files, e.g. the missing line to the above conversation would be c03253 +++$+++ Oh, come on! What are you talking about? The missing lines for the test dialogs always have c00000 as the conversation number but are otherwise formatted the same as the training file. While some of the short replies might be the same, every missing reply belongs to exactly one conversation. Task Your task is now to take the missing test replies and map them to the corresponding dialogs. More specifically you should write a script which can be called with the path to a file with the incomplete dialogs and the path to the missing replies and then outputs a file test_missing_with_predictions.txt in the same format as test_missing.txt only with actual conversation numbers from test_dialog.txt instead of c00000. You can chose whatever approach you want to solve the task, we only ask you to please write your code in Python 2.7 and if you use any external libraries provide a requirements.txt file from which these libraries can be installed with pip install -r requirements.txt (you might want to use a virtual environment and when you're done call pip freeze > requirements.txt). While it is okay to use other resources such as pretrained word embeddings to solve the task, we ask you not to train your algorithm using the original conversations provided with [1] as this would lead to overfitting, i.e. considered cheating. You should turn in all the code required to solve the task, i.e. which allows us to create the test_missing_with_predictions.txt file from the file without labels. Besides the accuracy of the predicted conversation labels we will also evaluate your code with respect to efficiency, maintainability, and readability (it might not hurt to have a look at some style guides). In addition to the code which solves the task please turn in a text file or pdf with answers to the following questions: Describe your approach. Which methods did you chose and why? How do you evaluate your performance? Where are the weaknesses of your approach? What has to be considered when applying an approach like this in practice?
Skills: Machine learning Artificial Intelligence Data mining Data Science
Hourly - Expert ($$$) - Est. Time: 1 to 3 months, Less than 10 hrs/week - Posted
Seeking ML expert to implement the following: (1) Port out current logic into TensorFlow. (2) Several Naive Bayes classifiers on a very large corpus. Please let us know your available hours per week and hourly rate, and an explanation of your previous projects that make you a good fit for these ML tasks. After that, we will follow up with an invite to interview. Thanks very much! Tyler
Skills: Machine learning Elasticsearch TensorFlow
Hourly - Intermediate ($$) - Est. Time: 1 to 3 months, 10-30 hrs/week - Posted
Looking for an individual with a background in biomedical engineering, electrical engineering, or physics who has a firm grasp manipulating and analyzing continuous time-series data. Working with the project manager, this project requires planning, developing, and implementing a physiological data analysis scheme comprising of signal processing and synchronization, feature engineering, and probabilistic classification analysis using machine learning methods in MATLAB and/or Python.
Skills: Machine learning Data Engineering MATLAB Python Numpy
Fixed-Price - Intermediate ($$) - Est. Budget: $25 - Posted
Hi, I am running a training institution on Analytics & Data Science. I am also planning to start few more courses on Hadoop & Big Data Analytics. I need someone who can design a Course Curriculum for Hadoop Developer & Big Data Analytics for me. Please Note: We are not providing Graduation/Masters degree, Hence the syllabus should not be too lengthy Please comeup with your experience and advice in order to make it best for students.
Skills: Machine learning Apache Hive Apache Spark Data Analytics
Fixed-Price - Intermediate ($$) - Est. Budget: $85 - Posted
data analytics (e.g., being able to look at how topics change over time). This position requires strong data analysis skills, including statistical analyses, and an interest in participating in the research process (e.g., thinking creatively, generating ideas). There are two types of files (because the link structure made it difficult to crawl all of the data at one time). They contain some overlapping information (title=postTitle user=postUser, numComments=numComments); you can see a link of each below with the field names below it. Here's a link to the July 2014 birth club: People would have joined this club around Nov 2013 and their children are now 2 years old. People who would be just finding out they're pregnant now would join the May 2017 birth club (which of course will have far fewer members and posts than the July 2014 one): July 2014 title, link, user, numHugs, numComments 2014 (this is all 12 months combined whereas the file above is just July) numComments, postDate, postEdited, postUser, postTitle, postText, postImg In case you're interested, here is overall number of posts per year when I crawled in March. It peaks around 2013 because BabyCenter got more popular over time, but the most recent boards haven't yet had time to grow. 2007: 51,582 2008: 185,020 2009: 825,787 2010: 1,185,906 2011: 1,255,787 2012: 1,620,140 2013: 1,900,360 2014: 1,575,572 2015: 1,252,975 2016: 446,300 Total: 10,299,429
Skills: Machine learning Python
Fixed-Price - Expert ($$$) - Est. Budget: $100 - Posted
Hi I am looking for MATLAB expert who can help out to change/modify/optimise my existing liver cancer segmenation code or give own suggestion to improve the accuracy upt till 95% with CT images dataset which is marked by the Radiologist . I have used mophological operations with trial and error determed threshold point . on top required some statistically performance evaluation for which I will provide you reference paper and formula
Skills: Machine learning C# MATLAB