tensorflow - My regression network is not learning -


i have data tested in simplest regression network tensorflow, , dont valid learning. here code:

import tensorflow tf  tf.variable_scope("inputs"):     tf_x = tf.placeholder(tf.float32, xs.shape, name="x")     tf_y = tf.placeholder(tf.float32, ys.shape, name="y")  tf.variable_scope("net"):     l1 = tf.layers.dense(tf_x, 3, tf.nn.relu)     weights = l1.graph.get_tensor_by_name(os.path.split(l1.name)[0] + "/kernel:0")     output = tf.layers.dense(l1, 1)     tf.summary.histogram("h_out", l1)     tf.summary.histogram("pred", output)     tf.summary.histogram("weights", weights)  loss = tf.losses.mean_squared_error(tf_y, output, scope="loss") optimizer = tf.train.gradientdescentoptimizer(learning_rate=0.05) train_op = optimizer.minimize(loss) tf.summary.scalar("loss", loss) sess = tf.session() sess.run(tf.global_variables_initializer())  writer = tf.summary.filewriter("./logs", sess.graph) merge_op = tf.summary.merge_all()  step in range(1000):     _, l, pred, result = sess.run([train_op, loss, output, merge_op], {tf_x:xs, tf_y:ys})     writer.add_summary(result, step)     if step % 100 == 0:         print (l)

my data :

x:         2.120      1.860      2.310      2.060      2.520      1.770      1.450      1.420      2.250      1.930      2.550      2.050      2.250      2.570      1.790      2.380      2.570      1.850      2.740      1.830      2.360      2.460   y:   27.00000   57.00000   98.00000  267.00000   59.00000  142.00000  110.00000  135.00000   91.00000  119.00000   62.00000   40.00000  166.00000  116.00000  335.00000   39.00000   67.00000   48.00000   35.00000   33.00000   48.00000   35.00000

here loss , weights:

loss weights


Comments

Post a Comment

Popular posts from this blog

javascript - Create a stacked percentage column -

i have collection of data : var data = [ {"p301a":"10","p301b":"7","p301c":"7","p301d":"3","p301e":"8","p301f":"8","p301g":"8","p301h":"8","p301i":"8","p301j":"8","p301k":"8","p301l":"8","p301m":"8","p301n":"8","p301o":"8","age":"31-40 years","profesion":"2","position":"2"}, {"p301a":"5","p301b":"4","p301c":"4","p301d":"4","p301e":"4","p301f":"4","p301g":"4","p301h":"4","p301i":"4","p301j":"4","p301k":"4...
Read more

Optimising Firebase database by automatically overwriting data -

in firebase app have following structure: posts latest all users can write posts > latest . want limit entries 20 posts save space in database. , posts beyond 20 unnecessary never shown on homepage. how can limit posts 20, when user writes posts > latest posts @ end drop off (be deleted) automatically? this sample should you. you need : const max_log_count = 20; exports.removeold = functions.database.ref('/posts/latest/{postid}').oncreate(event => { const parentref = event.data.ref.parent; return parentref.once('value').then(snapshot => { if (snapshot.numchildren() >= max_log_count) { let childcount = 0; const updates = {}; snapshot.foreach(function(child) { if (++childcount <= snapshot.numchildren() - max_log_count) { updates[child.key] = null; } }); // update parent. removes children. ...
Read more

javascript - Angular UI-Grid customTemplate directive causing rows to load slowly/? -

Image
each grid cell can have more 1 customtemplate loaded. have directive holds watcher on model_col_field listen cell updates on scroll when dom rows updated , recompile celltemplate . without watcher , $compile rows repeat on scroll, showing same set of rows. is there more angular way of approaching or way improve performance? relevant controller code: create gridoptions.data & gridoptions.columndefs : var customtemp = '<div ng-repeat="opt in grid.appscope.colvalueoptionsmodel" ng-init="tempstring=model_col_field.clazz+opt.key;"><comparison-directive dbo="model_col_field" attobj="opt" fff="grid.appscope.getattributetemplate(tempstring)" /></div>'; $scope.gridoptions.columndefs = [ { name:'yindex',pinnedleft:true,celltemplate: customtemp, displayname:'index'}, { name:'col1',celltemplate: customtemp}, { name:'col2',celltemplate: c...
Read more